Index: head/en_US.ISO8859-1/books/handbook/config/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/config/chapter.xml (revision 50807) +++ head/en_US.ISO8859-1/books/handbook/config/chapter.xml (revision 50808) @@ -1,3531 +1,3531 @@ Configuration and Tuning Chern Lee Written by Mike Smith Based on a tutorial written by Matt Dillon Also based on tuning(7) written by Synopsis system configuration system optimization One of the important aspects of &os; is proper system configuration. This chapter explains much of the &os; configuration process, including some of the parameters which can be set to tune a &os; system. After reading this chapter, you will know: The basics of rc.conf configuration and /usr/local/etc/rc.d startup scripts. How to configure and test a network card. How to configure virtual hosts on network devices. How to use the various configuration files in /etc. How to tune &os; using &man.sysctl.8; variables. How to tune disk performance and modify kernel limitations. Before reading this chapter, you should: Understand &unix; and &os; basics (). Be familiar with the basics of kernel configuration and compilation (). Starting Services Tom Rhodes Contributed by services Many users install third party software on &os; from the Ports Collection and require the installed services to be started upon system initialization. Services, such as mail/postfix or www/apache22 are just two of the many software packages which may be started during system initialization. This section explains the procedures available for starting third party software. In &os;, most included services, such as &man.cron.8;, are started through the system startup scripts. Extended Application Configuration Now that &os; includes rc.d, configuration of application startup is easier and provides more features. Using the key words discussed in , applications can be set to start after certain other services and extra flags can be passed through /etc/rc.conf in place of hard coded flags in the startup script. A basic script may look similar to the following: #!/bin/sh # # PROVIDE: utility # REQUIRE: DAEMON # KEYWORD: shutdown . /etc/rc.subr name=utility rcvar=utility_enable command="/usr/local/sbin/utility" load_rc_config $name # # DO NOT CHANGE THESE DEFAULT VALUES HERE # SET THEM IN THE /etc/rc.conf FILE # utility_enable=${utility_enable-"NO"} pidfile=${utility_pidfile-"/var/run/utility.pid"} run_rc_command "$1" This script will ensure that the provided utility will be started after the DAEMON pseudo-service. It also provides a method for setting and tracking the process ID (PID). This application could then have the following line placed in /etc/rc.conf: utility_enable="YES" This method allows for easier manipulation of command line arguments, inclusion of the default functions provided in /etc/rc.subr, compatibility with &man.rcorder.8;, and provides for easier configuration via rc.conf. Using Services to Start Services Other services can be started using &man.inetd.8;. Working with &man.inetd.8; and its configuration is described in depth in . In some cases, it may make more sense to use &man.cron.8; to start system services. This approach has a number of advantages as &man.cron.8; runs these processes as the owner of the &man.crontab.5;. This allows regular users to start and maintain their own applications. The @reboot feature of &man.cron.8;, may be used in place of the time specification. This causes the job to run when &man.cron.8; is started, normally during system initialization. Configuring &man.cron.8; Tom Rhodes Contributed by cron configuration One of the most useful utilities in &os; is cron. This utility runs in the background and regularly checks /etc/crontab for tasks to execute and searches /var/cron/tabs for custom crontab files. These files are used to schedule tasks which cron runs at the specified times. Each entry in a crontab defines a task to run and is known as a cron job. Two different types of configuration files are used: the system crontab, which should not be modified, and user crontabs, which can be created and edited as needed. The format used by these files is documented in &man.crontab.5;. The format of the system crontab, /etc/crontab includes a who column which does not exist in user crontabs. In the system crontab, cron runs the command as the user specified in this column. In a user crontab, all commands run as the user who created the crontab. User crontabs allow individual users to schedule their own tasks. The root user can also have a user crontab which can be used to schedule tasks that do not exist in the system crontab. Here is a sample entry from the system crontab, /etc/crontab: - # /etc/crontab - root's crontab for FreeBSD + # /etc/crontab - root's crontab for FreeBSD # -# $FreeBSD$ +# $FreeBSD$ # SHELL=/bin/sh PATH=/etc:/bin:/sbin:/usr/bin:/usr/sbin # #minute hour mday month wday who command # */5 * * * * root /usr/libexec/atrun Lines that begin with the # character are comments. A comment can be placed in the file as a reminder of what and why a desired action is performed. Comments cannot be on the same line as a command or else they will be interpreted as part of the command; they must be on a new line. Blank lines are ignored. The equals (=) character is used to define any environment settings. In this example, it is used to define the SHELL and PATH. If the SHELL is omitted, cron will use the default Bourne shell. If the PATH is omitted, the full path must be given to the command or script to run. This line defines the seven fields used in a system crontab: minute, hour, mday, month, wday, who, and command. The minute field is the time in minutes when the specified command will be run, the hour is the hour when the specified command will be run, the mday is the day of the month, month is the month, and wday is the day of the week. These fields must be numeric values, representing the twenty-four hour clock, or a *, representing all values for that field. The who field only exists in the system crontab and specifies which user the command should be run as. The last field is the command to be executed. This entry defines the values for this cron job. The */5, followed by several more * characters, specifies that /usr/libexec/atrun is invoked by root every five minutes of every hour, of every day and day of the week, of every month. Commands can include any number of switches. However, commands which extend to multiple lines need to be broken with the backslash \ continuation character. Creating a User Crontab To create a user crontab, invoke crontab in editor mode: &prompt.user; crontab -e This will open the user's crontab using the default text editor. The first time a user runs this command, it will open an empty file. Once a user creates a crontab, this command will open that file for editing. It is useful to add these lines to the top of the crontab file in order to set the environment variables and to remember the meanings of the fields in the crontab: SHELL=/bin/sh PATH=/etc:/bin:/sbin:/usr/bin:/usr/sbin # Order of crontab fields # minute hour mday month wday command Then add a line for each command or script to run, specifying the time to run the command. This example runs the specified custom Bourne shell script every day at two in the afternoon. Since the path to the script is not specified in PATH, the full path to the script is given: 0 14 * * * /usr/home/dru/bin/mycustomscript.sh Before using a custom script, make sure it is executable and test it with the limited set of environment variables set by cron. To replicate the environment that would be used to run the above cron entry, use: env -i SHELL=/bin/sh PATH=/etc:/bin:/sbin:/usr/bin:/usr/sbin HOME=/home/dru LOGNAME=dru /usr/home/dru/bin/mycustomscript.sh The environment set by cron is discussed in &man.crontab.5;. Checking that scripts operate correctly in a cron environment is especially important if they include any commands that delete files using wildcards. When finished editing the crontab, save the file. It will automatically be installed and cron will read the crontab and run its cron jobs at their specified times. To list the cron jobs in a crontab, use this command: &prompt.user; crontab -l 0 14 * * * /usr/home/dru/bin/mycustomscript.sh To remove all of the cron jobs in a user crontab: &prompt.user; crontab -r remove crontab for dru? y Managing Services in &os; Tom Rhodes Contributed by &os; uses the &man.rc.8; system of startup scripts during system initialization and for managing services. The scripts listed in /etc/rc.d provide basic services which can be controlled with the , , and options to &man.service.8;. For instance, &man.sshd.8; can be restarted with the following command: &prompt.root; service sshd restart This procedure can be used to start services on a running system. Services will be started automatically at boot time as specified in &man.rc.conf.5;. For example, to enable &man.natd.8; at system startup, add the following line to /etc/rc.conf: natd_enable="YES" If a line is already present, change the NO to YES. The &man.rc.8; scripts will automatically load any dependent services during the next boot, as described below. Since the &man.rc.8; system is primarily intended to start and stop services at system startup and shutdown time, the , and options will only perform their action if the appropriate /etc/rc.conf variable is set. For instance, sshd restart will only work if sshd_enable is set to in /etc/rc.conf. To , or a service regardless of the settings in /etc/rc.conf, these commands should be prefixed with one. For instance, to restart &man.sshd.8; regardless of the current /etc/rc.conf setting, execute the following command: &prompt.root; service sshd onerestart To check if a service is enabled in /etc/rc.conf, run the appropriate &man.rc.8; script with . This example checks to see if &man.sshd.8; is enabled in /etc/rc.conf: &prompt.root; service sshd rcvar # sshd # sshd_enable="YES" # (default: "") The # sshd line is output from the above command, not a root console. To determine whether or not a service is running, use . For instance, to verify that &man.sshd.8; is running: &prompt.root; service sshd status sshd is running as pid 433. In some cases, it is also possible to a service. This attempts to send a signal to an individual service, forcing the service to reload its configuration files. In most cases, this means sending the service a SIGHUP signal. Support for this feature is not included for every service. The &man.rc.8; system is used for network services and it also contributes to most of the system initialization. For instance, when the /etc/rc.d/bgfsck script is executed, it prints out the following message: Starting background file system checks in 60 seconds. This script is used for background file system checks, which occur only during system initialization. Many system services depend on other services to function properly. For example, &man.yp.8; and other RPC-based services may fail to start until after the &man.rpcbind.8; service has started. To resolve this issue, information about dependencies and other meta-data is included in the comments at the top of each startup script. The &man.rcorder.8; program is used to parse these comments during system initialization to determine the order in which system services should be invoked to satisfy the dependencies. The following key word must be included in all startup scripts as it is required by &man.rc.subr.8; to enable the startup script: PROVIDE: Specifies the services this file provides. The following key words may be included at the top of each startup script. They are not strictly necessary, but are useful as hints to &man.rcorder.8;: REQUIRE: Lists services which are required for this service. The script containing this key word will run after the specified services. BEFORE: Lists services which depend on this service. The script containing this key word will run before the specified services. By carefully setting these keywords for each startup script, an administrator has a fine-grained level of control of the startup order of the scripts, without the need for runlevels used by some &unix; operating systems. Additional information can be found in &man.rc.8; and &man.rc.subr.8;. Refer to this article for instructions on how to create custom &man.rc.8; scripts. Managing System-Specific Configuration rc files rc.conf The principal location for system configuration information is /etc/rc.conf. This file contains a wide range of configuration information and it is read at system startup to configure the system. It provides the configuration information for the rc* files. The entries in /etc/rc.conf override the default settings in /etc/defaults/rc.conf. The file containing the default settings should not be edited. Instead, all system-specific changes should be made to /etc/rc.conf. A number of strategies may be applied in clustered applications to separate site-wide configuration from system-specific configuration in order to reduce administration overhead. The recommended approach is to place system-specific configuration into /etc/rc.conf.local. For example, these entries in /etc/rc.conf apply to all systems: sshd_enable="YES" keyrate="fast" defaultrouter="10.1.1.254" Whereas these entries in /etc/rc.conf.local apply to this system only: hostname="node1.example.org" ifconfig_fxp0="inet 10.1.1.1/8" Distribute /etc/rc.conf to every system using an application such as rsync or puppet, while /etc/rc.conf.local remains unique. Upgrading the system will not overwrite /etc/rc.conf, so system configuration information will not be lost. Both /etc/rc.conf and /etc/rc.conf.local are parsed by &man.sh.1;. This allows system operators to create complex configuration scenarios. Refer to &man.rc.conf.5; for further information on this topic. Setting Up Network Interface Cards Marc Fonvieille Contributed by network cards configuration Adding and configuring a network interface card (NIC) is a common task for any &os; administrator. Locating the Correct Driver network cards driver First, determine the model of the NIC and the chip it uses. &os; supports a wide variety of NICs. Check the Hardware Compatibility List for the &os; release to see if the NIC is supported. If the NIC is supported, determine the name of the &os; driver for the NIC. Refer to /usr/src/sys/conf/NOTES and /usr/src/sys/arch/conf/NOTES for the list of NIC drivers with some information about the supported chipsets. When in doubt, read the manual page of the driver as it will provide more information about the supported hardware and any known limitations of the driver. The drivers for common NICs are already present in the GENERIC kernel, meaning the NIC should be probed during boot. The system's boot messages can be viewed by typing more /var/run/dmesg.boot and using the spacebar to scroll through the text. In this example, two Ethernet NICs using the &man.dc.4; driver are present on the system: dc0: <82c169 PNIC 10/100BaseTX> port 0xa000-0xa0ff mem 0xd3800000-0xd38 000ff irq 15 at device 11.0 on pci0 miibus0: <MII bus> on dc0 bmtphy0: <BCM5201 10/100baseTX PHY> PHY 1 on miibus0 bmtphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto dc0: Ethernet address: 00:a0:cc:da:da:da dc0: [ITHREAD] dc1: <82c169 PNIC 10/100BaseTX> port 0x9800-0x98ff mem 0xd3000000-0xd30 000ff irq 11 at device 12.0 on pci0 miibus1: <MII bus> on dc1 bmtphy1: <BCM5201 10/100baseTX PHY> PHY 1 on miibus1 bmtphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto dc1: Ethernet address: 00:a0:cc:da:da:db dc1: [ITHREAD] If the driver for the NIC is not present in GENERIC, but a driver is available, the driver will need to be loaded before the NIC can be configured and used. This may be accomplished in one of two ways: The easiest way is to load a kernel module for the NIC using &man.kldload.8;. To also automatically load the driver at boot time, add the appropriate line to /boot/loader.conf. Not all NIC drivers are available as modules. Alternatively, statically compile support for the NIC into a custom kernel. Refer to /usr/src/sys/conf/NOTES, /usr/src/sys/arch/conf/NOTES and the manual page of the driver to determine which line to add to the custom kernel configuration file. For more information about recompiling the kernel, refer to . If the NIC was detected at boot, the kernel does not need to be recompiled. Using &windows; <acronym>NDIS</acronym> Drivers NDIS NDISulator &windows; drivers µsoft.windows; device drivers KLD (kernel loadable object) Unfortunately, there are still many vendors that do not provide schematics for their drivers to the open source community because they regard such information as trade secrets. Consequently, the developers of &os; and other operating systems are left with two choices: develop the drivers by a long and pain-staking process of reverse engineering or using the existing driver binaries available for µsoft.windows; platforms. &os; provides native support for the Network Driver Interface Specification (NDIS). It includes &man.ndisgen.8; which can be used to convert a &windowsxp; driver into a format that can be used on &os;. Because the &man.ndis.4; driver uses a &windowsxp; binary, it only runs on &i386; and amd64 systems. PCI, CardBus, PCMCIA, and USB devices are supported. To use &man.ndisgen.8;, three things are needed: &os; kernel sources. A &windowsxp; driver binary with a .SYS extension. A &windowsxp; driver configuration file with a .INF extension. Download the .SYS and .INF files for the specific NIC. Generally, these can be found on the driver CD or at the vendor's website. The following examples use W32DRIVER.SYS and W32DRIVER.INF. The driver bit width must match the version of &os;. For &os;/i386, use a &windows; 32-bit driver. For &os;/amd64, a &windows; 64-bit driver is needed. The next step is to compile the driver binary into a loadable kernel module. As root, use &man.ndisgen.8;: &prompt.root; ndisgen /path/to/W32DRIVER.INF /path/to/W32DRIVER.SYS This command is interactive and prompts for any extra information it requires. A new kernel module will be generated in the current directory. Use &man.kldload.8; to load the new module: &prompt.root; kldload ./W32DRIVER_SYS.ko In addition to the generated kernel module, the ndis.ko and if_ndis.ko modules must be loaded. This should happen automatically when any module that depends on &man.ndis.4; is loaded. If not, load them manually, using the following commands: &prompt.root; kldload ndis &prompt.root; kldload if_ndis The first command loads the &man.ndis.4; miniport driver wrapper and the second loads the generated NIC driver. Check &man.dmesg.8; to see if there were any load errors. If all went well, the output should be similar to the following: ndis0: <Wireless-G PCI Adapter> mem 0xf4100000-0xf4101fff irq 3 at device 8.0 on pci1 ndis0: NDIS API version: 5.0 ndis0: Ethernet address: 0a:b1:2c:d3:4e:f5 ndis0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps ndis0: 11g rates: 6Mbps 9Mbps 12Mbps 18Mbps 36Mbps 48Mbps 54Mbps From here, ndis0 can be configured like any other NIC. To configure the system to load the &man.ndis.4; modules at boot time, copy the generated module, W32DRIVER_SYS.ko, to /boot/modules. Then, add the following line to /boot/loader.conf: W32DRIVER_SYS_load="YES" Configuring the Network Card network cards configuration Once the right driver is loaded for the NIC, the card needs to be configured. It may have been configured at installation time by &man.bsdinstall.8;. To display the NIC configuration, enter the following command: &prompt.user; ifconfig dc0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=80008<VLAN_MTU,LINKSTATE> ether 00:a0:cc:da:da:da inet 192.168.1.3 netmask 0xffffff00 broadcast 192.168.1.255 media: Ethernet autoselect (100baseTX <full-duplex>) status: active dc1: flags=8802<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=80008<VLAN_MTU,LINKSTATE> ether 00:a0:cc:da:da:db inet 10.0.0.1 netmask 0xffffff00 broadcast 10.0.0.255 media: Ethernet 10baseT/UTP status: no carrier lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> metric 0 mtu 16384 options=3<RXCSUM,TXCSUM> inet6 fe80::1%lo0 prefixlen 64 scopeid 0x4 inet6 ::1 prefixlen 128 inet 127.0.0.1 netmask 0xff000000 nd6 options=3<PERFORMNUD,ACCEPT_RTADV> In this example, the following devices were displayed: dc0: The first Ethernet interface. dc1: The second Ethernet interface. lo0: The loopback device. &os; uses the driver name followed by the order in which the card is detected at boot to name the NIC. For example, sis2 is the third NIC on the system using the &man.sis.4; driver. In this example, dc0 is up and running. The key indicators are: UP means that the card is configured and ready. The card has an Internet (inet) address, 192.168.1.3. It has a valid subnet mask (netmask), where 0xffffff00 is the same as 255.255.255.0. It has a valid broadcast address, 192.168.1.255. The MAC address of the card (ether) is 00:a0:cc:da:da:da. The physical media selection is on autoselection mode (media: Ethernet autoselect (100baseTX <full-duplex>)). In this example, dc1 is configured to run with 10baseT/UTP media. For more information on available media types for a driver, refer to its manual page. The status of the link (status) is active, indicating that the carrier signal is detected. For dc1, the status: no carrier status is normal when an Ethernet cable is not plugged into the card. If the &man.ifconfig.8; output had shown something similar to: dc0: flags=8843<BROADCAST,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=80008<VLAN_MTU,LINKSTATE> ether 00:a0:cc:da:da:da media: Ethernet autoselect (100baseTX <full-duplex>) status: active it would indicate the card has not been configured. The card must be configured as root. The NIC configuration can be performed from the command line with &man.ifconfig.8; but will not persist after a reboot unless the configuration is also added to /etc/rc.conf. If a DHCP server is present on the LAN, just add this line: ifconfig_dc0="DHCP" Replace dc0 with the correct value for the system. The line added, then, follow the instructions given in . If the network was configured during installation, some entries for the NIC(s) may be already present. Double check /etc/rc.conf before adding any lines. In the case, there is no DHCP server, the NIC(s) have to be configured manually. Add a line for each NIC present on the system, as seen in this example: ifconfig_dc0="inet 192.168.1.3 netmask 255.255.255.0" ifconfig_dc1="inet 10.0.0.1 netmask 255.255.255.0 media 10baseT/UTP" Replace dc0 and dc1 and the IP address information with the correct values for the system. Refer to the man page for the driver, &man.ifconfig.8;, and &man.rc.conf.5; for more details about the allowed options and the syntax of /etc/rc.conf. If the network is not using DNS, edit /etc/hosts to add the names and IP addresses of the hosts on the LAN, if they are not already there. For more information, refer to &man.hosts.5; and to /usr/share/examples/etc/hosts. If there is no DHCP server and access to the Internet is needed, manually configure the default gateway and the nameserver: &prompt.root; echo 'defaultrouter="your_default_router"' >> /etc/rc.conf &prompt.root; echo 'nameserver your_DNS_server' >> /etc/resolv.conf Testing and Troubleshooting Once the necessary changes to /etc/rc.conf are saved, a reboot can be used to test the network configuration and to verify that the system restarts without any configuration errors. Alternatively, apply the settings to the networking system with this command: &prompt.root; service netif restart If a default gateway has been set in /etc/rc.conf, also issue this command: &prompt.root; service routing restart Once the networking system has been relaunched, test the NICs. Testing the Ethernet Card network cards testing To verify that an Ethernet card is configured correctly, &man.ping.8; the interface itself, and then &man.ping.8; another machine on the LAN: &prompt.user; ping -c5 192.168.1.3 PING 192.168.1.3 (192.168.1.3): 56 data bytes 64 bytes from 192.168.1.3: icmp_seq=0 ttl=64 time=0.082 ms 64 bytes from 192.168.1.3: icmp_seq=1 ttl=64 time=0.074 ms 64 bytes from 192.168.1.3: icmp_seq=2 ttl=64 time=0.076 ms 64 bytes from 192.168.1.3: icmp_seq=3 ttl=64 time=0.108 ms 64 bytes from 192.168.1.3: icmp_seq=4 ttl=64 time=0.076 ms --- 192.168.1.3 ping statistics --- 5 packets transmitted, 5 packets received, 0% packet loss round-trip min/avg/max/stddev = 0.074/0.083/0.108/0.013 ms &prompt.user; ping -c5 192.168.1.2 PING 192.168.1.2 (192.168.1.2): 56 data bytes 64 bytes from 192.168.1.2: icmp_seq=0 ttl=64 time=0.726 ms 64 bytes from 192.168.1.2: icmp_seq=1 ttl=64 time=0.766 ms 64 bytes from 192.168.1.2: icmp_seq=2 ttl=64 time=0.700 ms 64 bytes from 192.168.1.2: icmp_seq=3 ttl=64 time=0.747 ms 64 bytes from 192.168.1.2: icmp_seq=4 ttl=64 time=0.704 ms --- 192.168.1.2 ping statistics --- 5 packets transmitted, 5 packets received, 0% packet loss round-trip min/avg/max/stddev = 0.700/0.729/0.766/0.025 ms To test network resolution, use the host name instead of the IP address. If there is no DNS server on the network, /etc/hosts must first be configured. To this purpose, edit /etc/hosts to add the names and IP addresses of the hosts on the LAN, if they are not already there. For more information, refer to &man.hosts.5; and to /usr/share/examples/etc/hosts. Troubleshooting network cards troubleshooting When troubleshooting hardware and software configurations, check the simple things first. Is the network cable plugged in? Are the network services properly configured? Is the firewall configured correctly? Is the NIC supported by &os;? Before sending a bug report, always check the Hardware Notes, update the version of &os; to the latest STABLE version, check the mailing list archives, and search the Internet. If the card works, yet performance is poor, read through &man.tuning.7;. Also, check the network configuration as incorrect network settings can cause slow connections. Some users experience one or two device timeout messages, which is normal for some cards. If they continue, or are bothersome, determine if the device is conflicting with another device. Double check the cable connections. Consider trying another card. To resolve watchdog timeout errors, first check the network cable. Many cards require a PCI slot which supports bus mastering. On some old motherboards, only one PCI slot allows it, usually slot 0. Check the NIC and the motherboard documentation to determine if that may be the problem. No route to host messages occur if the system is unable to route a packet to the destination host. This can happen if no default route is specified or if a cable is unplugged. Check the output of netstat -rn and make sure there is a valid route to the host. If there is not, read . ping: sendto: Permission denied error messages are often caused by a misconfigured firewall. If a firewall is enabled on &os; but no rules have been defined, the default policy is to deny all traffic, even &man.ping.8;. Refer to for more information. Sometimes performance of the card is poor or below average. In these cases, try setting the media selection mode from autoselect to the correct media selection. While this works for most hardware, it may or may not resolve the issue. Again, check all the network settings, and refer to &man.tuning.7;. Virtual Hosts virtual hosts IP aliases A common use of &os; is virtual site hosting, where one server appears to the network as many servers. This is achieved by assigning multiple network addresses to a single interface. A given network interface has one real address, and may have any number of alias addresses. These aliases are normally added by placing alias entries in /etc/rc.conf, as seen in this example: ifconfig_fxp0_alias0="inet xxx.xxx.xxx.xxx netmask xxx.xxx.xxx.xxx" Alias entries must start with alias0 using a sequential number such as alias0, alias1, and so on. The configuration process will stop at the first missing number. The calculation of alias netmasks is important. For a given interface, there must be one address which correctly represents the network's netmask. Any other addresses which fall within this network must have a netmask of all 1s, expressed as either 255.255.255.255 or 0xffffffff. For example, consider the case where the fxp0 interface is connected to two networks: 10.1.1.0 with a netmask of 255.255.255.0 and 202.0.75.16 with a netmask of 255.255.255.240. The system is to be configured to appear in the ranges 10.1.1.1 through 10.1.1.5 and 202.0.75.17 through 202.0.75.20. Only the first address in a given network range should have a real netmask. All the rest (10.1.1.2 through 10.1.1.5 and 202.0.75.18 through 202.0.75.20) must be configured with a netmask of 255.255.255.255. The following /etc/rc.conf entries configure the adapter correctly for this scenario: ifconfig_fxp0="inet 10.1.1.1 netmask 255.255.255.0" ifconfig_fxp0_alias0="inet 10.1.1.2 netmask 255.255.255.255" ifconfig_fxp0_alias1="inet 10.1.1.3 netmask 255.255.255.255" ifconfig_fxp0_alias2="inet 10.1.1.4 netmask 255.255.255.255" ifconfig_fxp0_alias3="inet 10.1.1.5 netmask 255.255.255.255" ifconfig_fxp0_alias4="inet 202.0.75.17 netmask 255.255.255.240" ifconfig_fxp0_alias5="inet 202.0.75.18 netmask 255.255.255.255" ifconfig_fxp0_alias6="inet 202.0.75.19 netmask 255.255.255.255" ifconfig_fxp0_alias7="inet 202.0.75.20 netmask 255.255.255.255" A simpler way to express this is with a space-separated list of IP address ranges. The first address will be given the indicated subnet mask and the additional addresses will have a subnet mask of 255.255.255.255. ifconfig_fxp0_aliases="inet 10.1.1.1-5/24 inet 202.0.75.17-20/28" Configuring System Logging Niclas Zeising Contributed by system logging syslog &man.syslogd.8; Generating and reading system logs is an important aspect of system administration. The information in system logs can be used to detect hardware and software issues as well as application and system configuration errors. This information also plays an important role in security auditing and incident response. Most system daemons and applications will generate log entries. &os; provides a system logger, syslogd, to manage logging. By default, syslogd is started when the system boots. This is controlled by the variable syslogd_enable in /etc/rc.conf. There are numerous application arguments that can be set using syslogd_flags in /etc/rc.conf. Refer to &man.syslogd.8; for more information on the available arguments. This section describes how to configure the &os; system logger for both local and remote logging and how to perform log rotation and log management. Configuring Local Logging syslog.conf The configuration file, /etc/syslog.conf, controls what syslogd does with log entries as they are received. There are several parameters to control the handling of incoming events. The facility describes which subsystem generated the message, such as the kernel or a daemon, and the level describes the severity of the event that occurred. This makes it possible to configure if and where a log message is logged, depending on the facility and level. It is also possible to take action depending on the application that sent the message, and in the case of remote logging, the hostname of the machine generating the logging event. This configuration file contains one line per action, where the syntax for each line is a selector field followed by an action field. The syntax of the selector field is facility.level which will match log messages from facility at level level or higher. It is also possible to add an optional comparison flag before the level to specify more precisely what is logged. Multiple selector fields can be used for the same action, and are separated with a semicolon (;). Using * will match everything. The action field denotes where to send the log message, such as to a file or remote log host. As an example, here is the default syslog.conf from &os;: - # $&os;$ + # $&os;$ # # Spaces ARE valid field separators in this file. However, # other *nix-like systems still insist on using tabs as field # separators. If you are sharing this file between systems, you # may want to use only tabs as field separators here. # Consult the syslog.conf(5) manpage. *.err;kern.warning;auth.notice;mail.crit /dev/console *.notice;authpriv.none;kern.debug;lpr.info;mail.crit;news.err /var/log/messages security.* /var/log/security auth.info;authpriv.info /var/log/auth.log mail.info /var/log/maillog lpr.info /var/log/lpd-errs ftp.info /var/log/xferlog cron.* /var/log/cron !-devd *.=debug /var/log/debug.log *.emerg * # uncomment this to log all writes to /dev/console to /var/log/console.log #console.info /var/log/console.log # uncomment this to enable logging of all log messages to /var/log/all.log # touch /var/log/all.log and chmod it to mode 600 before it will work #*.* /var/log/all.log # uncomment this to enable logging to a remote loghost named loghost #*.* @loghost # uncomment these if you're running inn # news.crit /var/log/news/news.crit # news.err /var/log/news/news.err # news.notice /var/log/news/news.notice # Uncomment this if you wish to see messages produced by devd # !devd # *.>=info !ppp *.* /var/log/ppp.log !* In this example: Line 8 matches all messages with a level of err or higher, as well as kern.warning, auth.notice and mail.crit, and sends these log messages to the console (/dev/console). Line 12 matches all messages from the mail facility at level info or above and logs the messages to /var/log/maillog. Line 17 uses a comparison flag (=) to only match messages at level debug and logs them to /var/log/debug.log. Line 33 is an example usage of a program specification. This makes the rules following it only valid for the specified program. In this case, only the messages generated by ppp are logged to /var/log/ppp.log. The available levels, in order from most to least critical are emerg, alert, crit, err, warning, notice, info, and debug. The facilities, in no particular order, are auth, authpriv, console, cron, daemon, ftp, kern, lpr, mail, mark, news, security, syslog, user, uucp, and local0 through local7. Be aware that other operating systems might have different facilities. To log everything of level notice and higher to /var/log/daemon.log, add the following entry: daemon.notice /var/log/daemon.log For more information about the different levels and facilities, refer to &man.syslog.3; and &man.syslogd.8;. For more information about /etc/syslog.conf, its syntax, and more advanced usage examples, see &man.syslog.conf.5;. Log Management and Rotation newsyslog newsyslog.conf log rotation log management Log files can grow quickly, taking up disk space and making it more difficult to locate useful information. Log management attempts to mitigate this. In &os;, newsyslog is used to manage log files. This built-in program periodically rotates and compresses log files, and optionally creates missing log files and signals programs when log files are moved. The log files may be generated by syslogd or by any other program which generates log files. While newsyslog is normally run from &man.cron.8;, it is not a system daemon. In the default configuration, it runs every hour. To know which actions to take, newsyslog reads its configuration file, /etc/newsyslog.conf. This file contains one line for each log file that newsyslog manages. Each line states the file owner, permissions, when to rotate that file, optional flags that affect log rotation, such as compression, and programs to signal when the log is rotated. Here is the default configuration in &os;: - # configuration file for newsyslog -# $FreeBSD$ + # configuration file for newsyslog +# $FreeBSD$ # # Entries which do not specify the '/pid_file' field will cause the # syslogd process to be signalled when that log file is rotated. This # action is only appropriate for log files which are written to by the # syslogd process (ie, files listed in /etc/syslog.conf). If there # is no process which needs to be signalled when a given log file is # rotated, then the entry for that file should include the 'N' flag. # # The 'flags' field is one or more of the letters: BCDGJNUXZ or a '-'. # # Note: some sites will want to select more restrictive protections than the # defaults. In particular, it may be desirable to switch many of the 644 # entries to 640 or 600. For example, some sites will consider the # contents of maillog, messages, and lpd-errs to be confidential. In the # future, these defaults may change to more conservative ones. # # logfilename [owner:group] mode count size when flags [/pid_file] [sig_num] /var/log/all.log 600 7 * @T00 J /var/log/amd.log 644 7 100 * J /var/log/auth.log 600 7 100 @0101T JC /var/log/console.log 600 5 100 * J /var/log/cron 600 3 100 * JC /var/log/daily.log 640 7 * @T00 JN /var/log/debug.log 600 7 100 * JC /var/log/kerberos.log 600 7 100 * J /var/log/lpd-errs 644 7 100 * JC /var/log/maillog 640 7 * @T00 JC /var/log/messages 644 5 100 @0101T JC /var/log/monthly.log 640 12 * $M1D0 JN /var/log/pflog 600 3 100 * JB /var/run/pflogd.pid /var/log/ppp.log root:network 640 3 100 * JC /var/log/devd.log 644 3 100 * JC /var/log/security 600 10 100 * JC /var/log/sendmail.st 640 10 * 168 B /var/log/utx.log 644 3 * @01T05 B /var/log/weekly.log 640 5 1 $W6D0 JN /var/log/xferlog 600 7 100 * JC Each line starts with the name of the log to be rotated, optionally followed by an owner and group for both rotated and newly created files. The mode field sets the permissions on the log file and count denotes how many rotated log files should be kept. The size and when fields tell newsyslog when to rotate the file. A log file is rotated when either its size is larger than the size field or when the time in the when field has passed. An asterisk (*) means that this field is ignored. The flags field gives further instructions, such as how to compress the rotated file or to create the log file if it is missing. The last two fields are optional and specify the name of the Process ID (PID) file of a process and a signal number to send to that process when the file is rotated. For more information on all fields, valid flags, and how to specify the rotation time, refer to &man.newsyslog.conf.5;. Since newsyslog is run from &man.cron.8;, it cannot rotate files more often than it is scheduled to run from &man.cron.8;. Configuring Remote Logging Tom Rhodes Contributed by Monitoring the log files of multiple hosts can become unwieldy as the number of systems increases. Configuring centralized logging can reduce some of the administrative burden of log file administration. In &os;, centralized log file aggregation, merging, and rotation can be configured using syslogd and newsyslog. This section demonstrates an example configuration, where host A, named logserv.example.com, will collect logging information for the local network. Host B, named logclient.example.com, will be configured to pass logging information to the logging server. Log Server Configuration A log server is a system that has been configured to accept logging information from other hosts. Before configuring a log server, check the following: If there is a firewall between the logging server and any logging clients, ensure that the firewall ruleset allows UDP port 514 for both the clients and the server. The logging server and all client machines must have forward and reverse entries in the local DNS. If the network does not have a DNS server, create entries in each system's /etc/hosts. Proper name resolution is required so that log entries are not rejected by the logging server. On the log server, edit /etc/syslog.conf to specify the name of the client to receive log entries from, the logging facility to be used, and the name of the log to store the host's log entries. This example adds the hostname of B, logs all facilities, and stores the log entries in /var/log/logclient.log. Sample Log Server Configuration +logclient.example.com *.* /var/log/logclient.log When adding multiple log clients, add a similar two-line entry for each client. More information about the available facilities may be found in &man.syslog.conf.5;. Next, configure /etc/rc.conf: syslogd_enable="YES" syslogd_flags="-a logclient.example.com -v -v" The first entry starts syslogd at system boot. The second entry allows log entries from the specified client. The increases the verbosity of logged messages. This is useful for tweaking facilities as administrators are able to see what type of messages are being logged under each facility. Multiple options may be specified to allow logging from multiple clients. IP addresses and whole netblocks may also be specified. Refer to &man.syslogd.8; for a full list of possible options. Finally, create the log file: &prompt.root; touch /var/log/logclient.log At this point, syslogd should be restarted and verified: &prompt.root; service syslogd restart &prompt.root; pgrep syslog If a PID is returned, the server restarted successfully, and client configuration can begin. If the server did not restart, consult /var/log/messages for the error. Log Client Configuration A logging client sends log entries to a logging server on the network. The client also keeps a local copy of its own logs. Once a logging server has been configured, edit /etc/rc.conf on the logging client: syslogd_enable="YES" syslogd_flags="-s -v -v" The first entry enables syslogd on boot up. The second entry prevents logs from being accepted by this client from other hosts () and increases the verbosity of logged messages. Next, define the logging server in the client's /etc/syslog.conf. In this example, all logged facilities are sent to a remote system, denoted by the @ symbol, with the specified hostname: *.* @logserv.example.com After saving the edit, restart syslogd for the changes to take effect: &prompt.root; service syslogd restart To test that log messages are being sent across the network, use &man.logger.1; on the client to send a message to syslogd: &prompt.root; logger "Test message from logclient" This message should now exist both in /var/log/messages on the client and /var/log/logclient.log on the log server. Debugging Log Servers If no messages are being received on the log server, the cause is most likely a network connectivity issue, a hostname resolution issue, or a typo in a configuration file. To isolate the cause, ensure that both the logging server and the logging client are able to ping each other using the hostname specified in their /etc/rc.conf. If this fails, check the network cabling, the firewall ruleset, and the hostname entries in the DNS server or /etc/hosts on both the logging server and clients. Repeat until the ping is successful from both hosts. If the ping succeeds on both hosts but log messages are still not being received, temporarily increase logging verbosity to narrow down the configuration issue. In the following example, /var/log/logclient.log on the logging server is empty and /var/log/messages on the logging client does not indicate a reason for the failure. To increase debugging output, edit the syslogd_flags entry on the logging server and issue a restart: syslogd_flags="-d -a logclient.example.com -v -v" &prompt.root; service syslogd restart Debugging data similar to the following will flash on the console immediately after the restart: logmsg: pri 56, flags 4, from logserv.example.com, msg syslogd: restart syslogd: restarted logmsg: pri 6, flags 4, from logserv.example.com, msg syslogd: kernel boot file is /boot/kernel/kernel Logging to FILE /var/log/messages syslogd: kernel boot file is /boot/kernel/kernel cvthname(192.168.1.10) validate: dgram from IP 192.168.1.10, port 514, name logclient.example.com; rejected in rule 0 due to name mismatch. In this example, the log messages are being rejected due to a typo which results in a hostname mismatch. The client's hostname should be logclient, not logclien. Fix the typo, issue a restart, and verify the results: &prompt.root; service syslogd restart logmsg: pri 56, flags 4, from logserv.example.com, msg syslogd: restart syslogd: restarted logmsg: pri 6, flags 4, from logserv.example.com, msg syslogd: kernel boot file is /boot/kernel/kernel syslogd: kernel boot file is /boot/kernel/kernel logmsg: pri 166, flags 17, from logserv.example.com, msg Dec 10 20:55:02 <syslog.err> logserv.example.com syslogd: exiting on signal 2 cvthname(192.168.1.10) validate: dgram from IP 192.168.1.10, port 514, name logclient.example.com; accepted in rule 0. logmsg: pri 15, flags 0, from logclient.example.com, msg Dec 11 02:01:28 trhodes: Test message 2 Logging to FILE /var/log/logclient.log Logging to FILE /var/log/messages At this point, the messages are being properly received and placed in the correct file. Security Considerations As with any network service, security requirements should be considered before implementing a logging server. Log files may contain sensitive data about services enabled on the local host, user accounts, and configuration data. Network data sent from the client to the server will not be encrypted or password protected. If a need for encryption exists, consider using security/stunnel, which will transmit the logging data over an encrypted tunnel. Local security is also an issue. Log files are not encrypted during use or after log rotation. Local users may access log files to gain additional insight into system configuration. Setting proper permissions on log files is critical. The built-in log rotator, newsyslog, supports setting permissions on newly created and rotated log files. Setting log files to mode 600 should prevent unwanted access by local users. Refer to &man.newsyslog.conf.5; for additional information. Configuration Files <filename>/etc</filename> Layout There are a number of directories in which configuration information is kept. These include: /etc Generic system-specific configuration information. /etc/defaults Default versions of system configuration files. /etc/mail Extra &man.sendmail.8; configuration and other MTA configuration files. /etc/ppp Configuration for both user- and kernel-ppp programs. /etc/namedb Default location for &man.named.8; data. Normally named.conf and zone files are stored here. /usr/local/etc Configuration files for installed applications. May contain per-application subdirectories. /usr/local/etc/rc.d &man.rc.8; scripts for installed applications. /var/db Automatically generated system-specific database files, such as the package database and the &man.locate.1; database. Hostnames hostname DNS <filename>/etc/resolv.conf</filename> resolv.conf How a &os; system accesses the Internet Domain Name System (DNS) is controlled by &man.resolv.conf.5;. The most common entries to /etc/resolv.conf are: nameserver The IP address of a name server the resolver should query. The servers are queried in the order listed with a maximum of three. search Search list for hostname lookup. This is normally determined by the domain of the local hostname. domain The local domain name. A typical /etc/resolv.conf looks like this: search example.com nameserver 147.11.1.11 nameserver 147.11.100.30 Only one of the search and domain options should be used. When using DHCP, &man.dhclient.8; usually rewrites /etc/resolv.conf with information received from the DHCP server. <filename>/etc/hosts</filename> hosts /etc/hosts is a simple text database which works in conjunction with DNS and NIS to provide host name to IP address mappings. Entries for local computers connected via a LAN can be added to this file for simplistic naming purposes instead of setting up a &man.named.8; server. Additionally, /etc/hosts can be used to provide a local record of Internet names, reducing the need to query external DNS servers for commonly accessed names. - # $&os;$ + # $&os;$ # # # Host Database # # This file should contain the addresses and aliases for local hosts that # share this file. Replace 'my.domain' below with the domainname of your # machine. # # In the presence of the domain name service or NIS, this file may # not be consulted at all; see /etc/nsswitch.conf for the resolution order. # # ::1 localhost localhost.my.domain 127.0.0.1 localhost localhost.my.domain # # Imaginary network. #10.0.0.2 myname.my.domain myname #10.0.0.3 myfriend.my.domain myfriend # # According to RFC 1918, you can use the following IP networks for # private nets which will never be connected to the Internet: # # 10.0.0.0 - 10.255.255.255 # 172.16.0.0 - 172.31.255.255 # 192.168.0.0 - 192.168.255.255 # # In case you want to be able to connect to the Internet, you need # real official assigned numbers. Do not try to invent your own network # numbers but instead get one from your network provider (if any) or # from your regional registry (ARIN, APNIC, LACNIC, RIPE NCC, or AfriNIC.) # The format of /etc/hosts is as follows: [Internet address] [official hostname] [alias1] [alias2] ... For example: 10.0.0.1 myRealHostname.example.com myRealHostname foobar1 foobar2 Consult &man.hosts.5; for more information. Tuning with &man.sysctl.8; sysctl tuning with sysctl &man.sysctl.8; is used to make changes to a running &os; system. This includes many advanced options of the TCP/IP stack and virtual memory system that can dramatically improve performance for an experienced system administrator. Over five hundred system variables can be read and set using &man.sysctl.8;. At its core, &man.sysctl.8; serves two functions: to read and to modify system settings. To view all readable variables: &prompt.user; sysctl -a To read a particular variable, specify its name: &prompt.user; sysctl kern.maxproc kern.maxproc: 1044 To set a particular variable, use the variable=value syntax: &prompt.root; sysctl kern.maxfiles=5000 kern.maxfiles: 2088 -> 5000 Settings of sysctl variables are usually either strings, numbers, or booleans, where a boolean is 1 for yes or 0 for no. To automatically set some variables each time the machine boots, add them to /etc/sysctl.conf. For more information, refer to &man.sysctl.conf.5; and . <filename>sysctl.conf</filename> sysctl.conf sysctl The configuration file for &man.sysctl.8;, /etc/sysctl.conf, looks much like /etc/rc.conf. Values are set in a variable=value form. The specified values are set after the system goes into multi-user mode. Not all variables are settable in this mode. For example, to turn off logging of fatal signal exits and prevent users from seeing processes started by other users, the following tunables can be set in /etc/sysctl.conf: # Do not log fatal signal exits (e.g., sig 11) kern.logsigexit=0 # Prevent users from seeing information about processes that # are being run under another UID. security.bsd.see_other_uids=0 &man.sysctl.8; Read-only Tom Rhodes Contributed by In some cases it may be desirable to modify read-only &man.sysctl.8; values, which will require a reboot of the system. For instance, on some laptop models the &man.cardbus.4; device will not probe memory ranges and will fail with errors similar to: cbb0: Could not map register memory device_probe_and_attach: cbb0 attach returned 12 The fix requires the modification of a read-only &man.sysctl.8; setting. Add to /boot/loader.conf and reboot. Now &man.cardbus.4; should work properly. Tuning Disks The following section will discuss various tuning mechanisms and options which may be applied to disk devices. In many cases, disks with mechanical parts, such as SCSI drives, will be the bottleneck driving down the overall system performance. While a solution is to install a drive without mechanical parts, such as a solid state drive, mechanical drives are not going away anytime in the near future. When tuning disks, it is advisable to utilize the features of the &man.iostat.8; command to test various changes to the system. This command will allow the user to obtain valuable information on system IO. Sysctl Variables <varname>vfs.vmiodirenable</varname> vfs.vmiodirenable The vfs.vmiodirenable &man.sysctl.8; variable may be set to either 0 (off) or 1 (on). It is set to 1 by default. This variable controls how directories are cached by the system. Most directories are small, using just a single fragment (typically 1 K) in the file system and typically 512 bytes in the buffer cache. With this variable turned off, the buffer cache will only cache a fixed number of directories, even if the system has a huge amount of memory. When turned on, this &man.sysctl.8; allows the buffer cache to use the VM page cache to cache the directories, making all the memory available for caching directories. However, the minimum in-core memory used to cache a directory is the physical page size (typically 4 K) rather than 512  bytes. Keeping this option enabled is recommended if the system is running any services which manipulate large numbers of files. Such services can include web caches, large mail systems, and news systems. Keeping this option on will generally not reduce performance, even with the wasted memory, but one should experiment to find out. <varname>vfs.write_behind</varname> vfs.write_behind The vfs.write_behind &man.sysctl.8; variable defaults to 1 (on). This tells the file system to issue media writes as full clusters are collected, which typically occurs when writing large sequential files. This avoids saturating the buffer cache with dirty buffers when it would not benefit I/O performance. However, this may stall processes and under certain circumstances should be turned off. <varname>vfs.hirunningspace</varname> vfs.hirunningspace The vfs.hirunningspace &man.sysctl.8; variable determines how much outstanding write I/O may be queued to disk controllers system-wide at any given instance. The default is usually sufficient, but on machines with many disks, try bumping it up to four or five megabytes. Setting too high a value which exceeds the buffer cache's write threshold can lead to bad clustering performance. Do not set this value arbitrarily high as higher write values may add latency to reads occurring at the same time. There are various other buffer cache and VM page cache related &man.sysctl.8; values. Modifying these values is not recommended as the VM system does a good job of automatically tuning itself. <varname>vm.swap_idle_enabled</varname> vm.swap_idle_enabled The vm.swap_idle_enabled &man.sysctl.8; variable is useful in large multi-user systems with many active login users and lots of idle processes. Such systems tend to generate continuous pressure on free memory reserves. Turning this feature on and tweaking the swapout hysteresis (in idle seconds) via vm.swap_idle_threshold1 and vm.swap_idle_threshold2 depresses the priority of memory pages associated with idle processes more quickly then the normal pageout algorithm. This gives a helping hand to the pageout daemon. Only turn this option on if needed, because the tradeoff is essentially pre-page memory sooner rather than later which eats more swap and disk bandwidth. In a small system this option will have a determinable effect, but in a large system that is already doing moderate paging, this option allows the VM system to stage whole processes into and out of memory easily. <varname>hw.ata.wc</varname> hw.ata.wc Turning off IDE write caching reduces write bandwidth to IDE disks, but may sometimes be necessary due to data consistency issues introduced by hard drive vendors. The problem is that some IDE drives lie about when a write completes. With IDE write caching turned on, IDE hard drives write data to disk out of order and will sometimes delay writing some blocks indefinitely when under heavy disk load. A crash or power failure may cause serious file system corruption. Check the default on the system by observing the hw.ata.wc &man.sysctl.8; variable. If IDE write caching is turned off, one can set this read-only variable to 1 in /boot/loader.conf in order to enable it at boot time. For more information, refer to &man.ata.4;. <literal>SCSI_DELAY</literal> (<varname>kern.cam.scsi_delay</varname>) kern.cam.scsi_delay kernel options SCSI DELAY The SCSI_DELAY kernel configuration option may be used to reduce system boot times. The defaults are fairly high and can be responsible for 15 seconds of delay in the boot process. Reducing it to 5 seconds usually works with modern drives. The kern.cam.scsi_delay boot time tunable should be used. The tunable and kernel configuration option accept values in terms of milliseconds and not seconds. Soft Updates Soft Updates &man.tunefs.8; To fine-tune a file system, use &man.tunefs.8;. This program has many different options. To toggle Soft Updates on and off, use: &prompt.root; tunefs -n enable /filesystem &prompt.root; tunefs -n disable /filesystem A file system cannot be modified with &man.tunefs.8; while it is mounted. A good time to enable Soft Updates is before any partitions have been mounted, in single-user mode. Soft Updates is recommended for UFS file systems as it drastically improves meta-data performance, mainly file creation and deletion, through the use of a memory cache. There are two downsides to Soft Updates to be aware of. First, Soft Updates guarantee file system consistency in the case of a crash, but could easily be several seconds or even a minute behind updating the physical disk. If the system crashes, unwritten data may be lost. Secondly, Soft Updates delay the freeing of file system blocks. If the root file system is almost full, performing a major update, such as make installworld, can cause the file system to run out of space and the update to fail. More Details About Soft Updates Soft Updates details Meta-data updates are updates to non-content data like inodes or directories. There are two traditional approaches to writing a file system's meta-data back to disk. Historically, the default behavior was to write out meta-data updates synchronously. If a directory changed, the system waited until the change was actually written to disk. The file data buffers (file contents) were passed through the buffer cache and backed up to disk later on asynchronously. The advantage of this implementation is that it operates safely. If there is a failure during an update, meta-data is always in a consistent state. A file is either created completely or not at all. If the data blocks of a file did not find their way out of the buffer cache onto the disk by the time of the crash, &man.fsck.8; recognizes this and repairs the file system by setting the file length to 0. Additionally, the implementation is clear and simple. The disadvantage is that meta-data changes are slow. For example, rm -r touches all the files in a directory sequentially, but each directory change will be written synchronously to the disk. This includes updates to the directory itself, to the inode table, and possibly to indirect blocks allocated by the file. Similar considerations apply for unrolling large hierarchies using tar -x. The second approach is to use asynchronous meta-data updates. This is the default for a UFS file system mounted with mount -o async. Since all meta-data updates are also passed through the buffer cache, they will be intermixed with the updates of the file content data. The advantage of this implementation is there is no need to wait until each meta-data update has been written to disk, so all operations which cause huge amounts of meta-data updates work much faster than in the synchronous case. This implementation is still clear and simple, so there is a low risk for bugs creeping into the code. The disadvantage is that there is no guarantee for a consistent state of the file system. If there is a failure during an operation that updated large amounts of meta-data, like a power failure or someone pressing the reset button, the file system will be left in an unpredictable state. There is no opportunity to examine the state of the file system when the system comes up again as the data blocks of a file could already have been written to the disk while the updates of the inode table or the associated directory were not. It is impossible to implement a &man.fsck.8; which is able to clean up the resulting chaos because the necessary information is not available on the disk. If the file system has been damaged beyond repair, the only choice is to reformat it and restore from backup. The usual solution for this problem is to implement dirty region logging, which is also referred to as journaling. Meta-data updates are still written synchronously, but only into a small region of the disk. Later on, they are moved to their proper location. Because the logging area is a small, contiguous region on the disk, there are no long distances for the disk heads to move, even during heavy operations, so these operations are quicker than synchronous updates. Additionally, the complexity of the implementation is limited, so the risk of bugs being present is low. A disadvantage is that all meta-data is written twice, once into the logging region and once to the proper location, so performance pessimization might result. On the other hand, in case of a crash, all pending meta-data operations can be either quickly rolled back or completed from the logging area after the system comes up again, resulting in a fast file system startup. Kirk McKusick, the developer of Berkeley FFS, solved this problem with Soft Updates. All pending meta-data updates are kept in memory and written out to disk in a sorted sequence (ordered meta-data updates). This has the effect that, in case of heavy meta-data operations, later updates to an item catch the earlier ones which are still in memory and have not already been written to disk. All operations are generally performed in memory before the update is written to disk and the data blocks are sorted according to their position so that they will not be on the disk ahead of their meta-data. If the system crashes, an implicit log rewind causes all operations which were not written to the disk appear as if they never happened. A consistent file system state is maintained that appears to be the one of 30 to 60 seconds earlier. The algorithm used guarantees that all resources in use are marked as such in their blocks and inodes. After a crash, the only resource allocation error that occurs is that resources are marked as used which are actually free. &man.fsck.8; recognizes this situation, and frees the resources that are no longer used. It is safe to ignore the dirty state of the file system after a crash by forcibly mounting it with mount -f. In order to free resources that may be unused, &man.fsck.8; needs to be run at a later time. This is the idea behind the background &man.fsck.8;: at system startup time, only a snapshot of the file system is recorded and &man.fsck.8; is run afterwards. All file systems can then be mounted dirty, so the system startup proceeds in multi-user mode. Then, background &man.fsck.8; is scheduled for all file systems where this is required, to free resources that may be unused. File systems that do not use Soft Updates still need the usual foreground &man.fsck.8;. The advantage is that meta-data operations are nearly as fast as asynchronous updates and are faster than logging, which has to write the meta-data twice. The disadvantages are the complexity of the code, a higher memory consumption, and some idiosyncrasies. After a crash, the state of the file system appears to be somewhat older. In situations where the standard synchronous approach would have caused some zero-length files to remain after the &man.fsck.8;, these files do not exist at all with Soft Updates because neither the meta-data nor the file contents have been written to disk. Disk space is not released until the updates have been written to disk, which may take place some time after running &man.rm.1;. This may cause problems when installing large amounts of data on a file system that does not have enough free space to hold all the files twice. Tuning Kernel Limits tuning kernel limits File/Process Limits <varname>kern.maxfiles</varname> kern.maxfiles The kern.maxfiles &man.sysctl.8; variable can be raised or lowered based upon system requirements. This variable indicates the maximum number of file descriptors on the system. When the file descriptor table is full, file: table is full will show up repeatedly in the system message buffer, which can be viewed using &man.dmesg.8;. Each open file, socket, or fifo uses one file descriptor. A large-scale production server may easily require many thousands of file descriptors, depending on the kind and number of services running concurrently. In older &os; releases, the default value of kern.maxfiles is derived from in the kernel configuration file. kern.maxfiles grows proportionally to the value of . When compiling a custom kernel, consider setting this kernel configuration option according to the use of the system. From this number, the kernel is given most of its pre-defined limits. Even though a production machine may not have 256 concurrent users, the resources needed may be similar to a high-scale web server. The read-only &man.sysctl.8; variable kern.maxusers is automatically sized at boot based on the amount of memory available in the system, and may be determined at run-time by inspecting the value of kern.maxusers. Some systems require larger or smaller values of kern.maxusers and values of 64, 128, and 256 are not uncommon. Going above 256 is not recommended unless a huge number of file descriptors is needed. Many of the tunable values set to their defaults by kern.maxusers may be individually overridden at boot-time or run-time in /boot/loader.conf. Refer to &man.loader.conf.5; and /boot/defaults/loader.conf for more details and some hints. In older releases, the system will auto-tune maxusers if it is set to 0. The auto-tuning algorithm sets maxusers equal to the amount of memory in the system, with a minimum of 32, and a maximum of 384.. When setting this option, set maxusers to at least 4, especially if the system runs &xorg; or is used to compile software. The most important table set by maxusers is the maximum number of processes, which is set to 20 + 16 * maxusers. If maxusers is set to 1, there can only be 36 simultaneous processes, including the 18 or so that the system starts up at boot time and the 15 or so used by &xorg;. Even a simple task like reading a manual page will start up nine processes to filter, decompress, and view it. Setting maxusers to 64 allows up to 1044 simultaneous processes, which should be enough for nearly all uses. If, however, the proc table full error is displayed when trying to start another program, or a server is running with a large number of simultaneous users, increase the number and rebuild. maxusers does not limit the number of users which can log into the machine. It instead sets various table sizes to reasonable values considering the maximum number of users on the system and how many processes each user will be running. <varname>kern.ipc.soacceptqueue</varname> kern.ipc.soacceptqueue The kern.ipc.soacceptqueue &man.sysctl.8; variable limits the size of the listen queue for accepting new TCP connections. The default value of 128 is typically too low for robust handling of new connections on a heavily loaded web server. For such environments, it is recommended to increase this value to 1024 or higher. A service such as &man.sendmail.8;, or Apache may itself limit the listen queue size, but will often have a directive in its configuration file to adjust the queue size. Large listen queues do a better job of avoiding Denial of Service (DoS) attacks. Network Limits The NMBCLUSTERS kernel configuration option dictates the amount of network Mbufs available to the system. A heavily-trafficked server with a low number of Mbufs will hinder performance. Each cluster represents approximately 2 K of memory, so a value of 1024 represents 2 megabytes of kernel memory reserved for network buffers. A simple calculation can be done to figure out how many are needed. A web server which maxes out at 1000 simultaneous connections where each connection uses a 6 K receive and 16 K send buffer, requires approximately 32 MB worth of network buffers to cover the web server. A good rule of thumb is to multiply by 2, so 2x32 MB / 2 KB = 64 MB / 2 kB = 32768. Values between 4096 and 32768 are recommended for machines with greater amounts of memory. Never specify an arbitrarily high value for this parameter as it could lead to a boot time crash. To observe network cluster usage, use with &man.netstat.1;. The kern.ipc.nmbclusters loader tunable should be used to tune this at boot time. Only older versions of &os; will require the use of the NMBCLUSTERS kernel &man.config.8; option. For busy servers that make extensive use of the &man.sendfile.2; system call, it may be necessary to increase the number of &man.sendfile.2; buffers via the NSFBUFS kernel configuration option or by setting its value in /boot/loader.conf (see &man.loader.8; for details). A common indicator that this parameter needs to be adjusted is when processes are seen in the sfbufa state. The &man.sysctl.8; variable kern.ipc.nsfbufs is read-only. This parameter nominally scales with kern.maxusers, however it may be necessary to tune accordingly. Even though a socket has been marked as non-blocking, calling &man.sendfile.2; on the non-blocking socket may result in the &man.sendfile.2; call blocking until enough struct sf_buf's are made available. <varname>net.inet.ip.portrange.*</varname> net.inet.ip.portrange.* The net.inet.ip.portrange.* &man.sysctl.8; variables control the port number ranges automatically bound to TCP and UDP sockets. There are three ranges: a low range, a default range, and a high range. Most network programs use the default range which is controlled by net.inet.ip.portrange.first and net.inet.ip.portrange.last, which default to 1024 and 5000, respectively. Bound port ranges are used for outgoing connections and it is possible to run the system out of ports under certain circumstances. This most commonly occurs when running a heavily loaded web proxy. The port range is not an issue when running a server which handles mainly incoming connections, such as a web server, or has a limited number of outgoing connections, such as a mail relay. For situations where there is a shortage of ports, it is recommended to increase net.inet.ip.portrange.last modestly. A value of 10000, 20000 or 30000 may be reasonable. Consider firewall effects when changing the port range. Some firewalls may block large ranges of ports, usually low-numbered ports, and expect systems to use higher ranges of ports for outgoing connections. For this reason, it is not recommended that the value of net.inet.ip.portrange.first be lowered. <literal>TCP</literal> Bandwidth Delay Product TCP Bandwidth Delay Product Limiting net.inet.tcp.inflight.enable TCP bandwidth delay product limiting can be enabled by setting the net.inet.tcp.inflight.enable &man.sysctl.8; variable to 1. This instructs the system to attempt to calculate the bandwidth delay product for each connection and limit the amount of data queued to the network to just the amount required to maintain optimum throughput. This feature is useful when serving data over modems, Gigabit Ethernet, high speed WAN links, or any other link with a high bandwidth delay product, especially when also using window scaling or when a large send window has been configured. When enabling this option, also set net.inet.tcp.inflight.debug to 0 to disable debugging. For production use, setting net.inet.tcp.inflight.min to at least 6144 may be beneficial. Setting high minimums may effectively disable bandwidth limiting, depending on the link. The limiting feature reduces the amount of data built up in intermediate route and switch packet queues and reduces the amount of data built up in the local host's interface queue. With fewer queued packets, interactive connections, especially over slow modems, will operate with lower Round Trip Times. This feature only effects server side data transmission such as uploading. It has no effect on data reception or downloading. Adjusting net.inet.tcp.inflight.stab is not recommended. This parameter defaults to 20, representing 2 maximal packets added to the bandwidth delay product window calculation. The additional window is required to stabilize the algorithm and improve responsiveness to changing conditions, but it can also result in higher &man.ping.8; times over slow links, though still much lower than without the inflight algorithm. In such cases, try reducing this parameter to 15, 10, or 5 and reducing net.inet.tcp.inflight.min to a value such as 3500 to get the desired effect. Reducing these parameters should be done as a last resort only. Virtual Memory <varname>kern.maxvnodes</varname> A vnode is the internal representation of a file or directory. Increasing the number of vnodes available to the operating system reduces disk I/O. Normally, this is handled by the operating system and does not need to be changed. In some cases where disk I/O is a bottleneck and the system is running out of vnodes, this setting needs to be increased. The amount of inactive and free RAM will need to be taken into account. To see the current number of vnodes in use: &prompt.root; sysctl vfs.numvnodes vfs.numvnodes: 91349 To see the maximum vnodes: &prompt.root; sysctl kern.maxvnodes kern.maxvnodes: 100000 If the current vnode usage is near the maximum, try increasing kern.maxvnodes by a value of 1000. Keep an eye on the number of vfs.numvnodes. If it climbs up to the maximum again, kern.maxvnodes will need to be increased further. Otherwise, a shift in memory usage as reported by &man.top.1; should be visible and more memory should be active. Adding Swap Space Sometimes a system requires more swap space. This section describes two methods to increase swap space: adding swap to an existing partition or new hard drive, and creating a swap file on an existing partition. For information on how to encrypt swap space, which options exist, and why it should be done, refer to . Swap on a New Hard Drive or Existing Partition Adding a new hard drive for swap gives better performance than using a partition on an existing drive. Setting up partitions and hard drives is explained in while discusses partition layouts and swap partition size considerations. Use swapon to add a swap partition to the system. For example: &prompt.root; swapon /dev/ada1s1b It is possible to use any partition not currently mounted, even if it already contains data. Using swapon on a partition that contains data will overwrite and destroy that data. Make sure that the partition to be added as swap is really the intended partition before running swapon. To automatically add this swap partition on boot, add an entry to /etc/fstab: /dev/ada1s1b none swap sw 0 0 See &man.fstab.5; for an explanation of the entries in /etc/fstab. More information about swapon can be found in &man.swapon.8;. Creating a Swap File These examples create a 64M swap file called /usr/swap0 instead of using a partition. Using swap files requires that the module needed by &man.md.4; has either been built into the kernel or has been loaded before swap is enabled. See for information about building a custom kernel. Creating a Swap File on &os; 10.<replaceable>X</replaceable> and Later Create the swap file: &prompt.root; dd if=/dev/zero of=/usr/swap0 bs=1m count=64 Set the proper permissions on the new file: &prompt.root; chmod 0600 /usr/swap0 Inform the system about the swap file by adding a line to /etc/fstab: md99 none swap sw,file=/usr/swap0,late 0 0 The &man.md.4; device md99 is used, leaving lower device numbers available for interactive use. Swap space will be added on system startup. To add swap space immediately, use &man.swapon.8;: &prompt.root; swapon -aL Creating a Swap File on &os; 9.<replaceable>X</replaceable> and Earlier Create the swap file, /usr/swap0: &prompt.root; dd if=/dev/zero of=/usr/swap0 bs=1m count=64 Set the proper permissions on /usr/swap0: &prompt.root; chmod 0600 /usr/swap0 Enable the swap file in /etc/rc.conf: swapfile="/usr/swap0" # Set to name of swap file Swap space will be added on system startup. To enable the swap file immediately, specify a free memory device. Refer to for more information about memory devices. &prompt.root; mdconfig -a -t vnode -f /usr/swap0 -u 0 && swapon /dev/md0 Power and Resource Management Hiten Pandya Written by Tom Rhodes It is important to utilize hardware resources in an efficient manner. Power and resource management allows the operating system to monitor system limits and to possibly provide an alert if the system temperature increases unexpectedly. An early specification for providing power management was the Advanced Power Management (APM) facility. APM controls the power usage of a system based on its activity. However, it was difficult and inflexible for operating systems to manage the power usage and thermal properties of a system. The hardware was managed by the BIOS and the user had limited configurability and visibility into the power management settings. The APM BIOS is supplied by the vendor and is specific to the hardware platform. An APM driver in the operating system mediates access to the APM Software Interface, which allows management of power levels. There are four major problems in APM. First, power management is done by the vendor-specific BIOS, separate from the operating system. For example, the user can set idle-time values for a hard drive in the APM BIOS so that, when exceeded, the BIOS spins down the hard drive without the consent of the operating system. Second, the APM logic is embedded in the BIOS, and it operates outside the scope of the operating system. This means that users can only fix problems in the APM BIOS by flashing a new one into the ROM, which is a dangerous procedure with the potential to leave the system in an unrecoverable state if it fails. Third, APM is a vendor-specific technology, meaning that there is a lot of duplication of efforts and bugs found in one vendor's BIOS may not be solved in others. Lastly, the APM BIOS did not have enough room to implement a sophisticated power policy or one that can adapt well to the purpose of the machine. The Plug and Play BIOS (PNPBIOS) was unreliable in many situations. PNPBIOS is 16-bit technology, so the operating system has to use 16-bit emulation in order to interface with PNPBIOS methods. &os; provides an APM driver as APM should still be used for systems manufactured at or before the year 2000. The driver is documented in &man.apm.4;. ACPI APM The successor to APM is the Advanced Configuration and Power Interface (ACPI). ACPI is a standard written by an alliance of vendors to provide an interface for hardware resources and power management. It is a key element in Operating System-directed configuration and Power Management as it provides more control and flexibility to the operating system. This chapter demonstrates how to configure ACPI on &os;. It then offers some tips on how to debug ACPI and how to submit a problem report containing debugging information so that developers can diagnosis and fix ACPI issues. Configuring <acronym>ACPI</acronym> In &os; the &man.acpi.4; driver is loaded by default at system boot and should not be compiled into the kernel. This driver cannot be unloaded after boot because the system bus uses it for various hardware interactions. However, if the system is experiencing problems, ACPI can be disabled altogether by rebooting after setting hint.acpi.0.disabled="1" in /boot/loader.conf or by setting this variable at the loader prompt, as described in . ACPI and APM cannot coexist and should be used separately. The last one to load will terminate if the driver notices the other is running. ACPI can be used to put the system into a sleep mode with acpiconf, the flag, and a number from 1 to 5. Most users only need 1 (quick suspend to RAM) or 3 (suspend to RAM). Option 5 performs a soft-off which is the same as running halt -p. Other options are available using sysctl. Refer to &man.acpi.4; and &man.acpiconf.8; for more information. Common Problems ACPI ACPI is present in all modern computers that conform to the ia32 (x86), ia64 (Itanium), and amd64 (AMD) architectures. The full standard has many features including CPU performance management, power planes control, thermal zones, various battery systems, embedded controllers, and bus enumeration. Most systems implement less than the full standard. For instance, a desktop system usually only implements bus enumeration while a laptop might have cooling and battery management support as well. Laptops also have suspend and resume, with their own associated complexity. An ACPI-compliant system has various components. The BIOS and chipset vendors provide various fixed tables, such as FADT, in memory that specify things like the APIC map (used for SMP), config registers, and simple configuration values. Additionally, a bytecode table, the Differentiated System Description Table DSDT, specifies a tree-like name space of devices and methods. The ACPI driver must parse the fixed tables, implement an interpreter for the bytecode, and modify device drivers and the kernel to accept information from the ACPI subsystem. For &os;, &intel; has provided an interpreter (ACPI-CA) that is shared with &linux; and NetBSD. The path to the ACPI-CA source code is src/sys/contrib/dev/acpica. The glue code that allows ACPI-CA to work on &os; is in src/sys/dev/acpica/Osd. Finally, drivers that implement various ACPI devices are found in src/sys/dev/acpica. ACPI problems For ACPI to work correctly, all the parts have to work correctly. Here are some common problems, in order of frequency of appearance, and some possible workarounds or fixes. If a fix does not resolve the issue, refer to for instructions on how to submit a bug report. Mouse Issues In some cases, resuming from a suspend operation will cause the mouse to fail. A known work around is to add hint.psm.0.flags="0x3000" to /boot/loader.conf. Suspend/Resume ACPI has three suspend to RAM (STR) states, S1-S3, and one suspend to disk state (STD), called S4. STD can be implemented in two separate ways. The S4BIOS is a BIOS-assisted suspend to disk and S4OS is implemented entirely by the operating system. The normal state the system is in when plugged in but not powered up is soft off (S5). Use sysctl hw.acpi to check for the suspend-related items. These example results are from a Thinkpad: hw.acpi.supported_sleep_state: S3 S4 S5 hw.acpi.s4bios: 0 Use acpiconf -s to test S3, S4, and S5. An of one (1) indicates S4BIOS support instead of S4 operating system support. When testing suspend/resume, start with S1, if supported. This state is most likely to work since it does not require much driver support. No one has implemented S2, which is similar to S1. Next, try S3. This is the deepest STR state and requires a lot of driver support to properly reinitialize the hardware. A common problem with suspend/resume is that many device drivers do not save, restore, or reinitialize their firmware, registers, or device memory properly. As a first attempt at debugging the problem, try: &prompt.root; sysctl debug.bootverbose=1 &prompt.root; sysctl debug.acpi.suspend_bounce=1 &prompt.root; acpiconf -s 3 This test emulates the suspend/resume cycle of all device drivers without actually going into S3 state. In some cases, problems such as losing firmware state, device watchdog time out, and retrying forever, can be captured with this method. Note that the system will not really enter S3 state, which means devices may not lose power, and many will work fine even if suspend/resume methods are totally missing, unlike real S3 state. Harder cases require additional hardware, such as a serial port and cable for debugging through a serial console, a Firewire port and cable for using &man.dcons.4;, and kernel debugging skills. To help isolate the problem, unload as many drivers as possible. If it works, narrow down which driver is the problem by loading drivers until it fails again. Typically, binary drivers like nvidia.ko, display drivers, and USB will have the most problems while Ethernet interfaces usually work fine. If drivers can be properly loaded and unloaded, automate this by putting the appropriate commands in /etc/rc.suspend and /etc/rc.resume. Try setting to 1 if the display is messed up after resume. Try setting longer or shorter values for to see if that helps. Try loading a recent &linux; distribution to see if suspend/resume works on the same hardware. If it works on &linux;, it is likely a &os; driver problem. Narrowing down which driver causes the problem will assist developers in fixing the problem. Since the ACPI maintainers rarely maintain other drivers, such as sound or ATA, any driver problems should also be posted to the &a.current.name; list and mailed to the driver maintainer. Advanced users can include debugging &man.printf.3;s in a problematic driver to track down where in its resume function it hangs. Finally, try disabling ACPI and enabling APM instead. If suspend/resume works with APM, stick with APM, especially on older hardware (pre-2000). It took vendors a while to get ACPI support correct and older hardware is more likely to have BIOS problems with ACPI. System Hangs Most system hangs are a result of lost interrupts or an interrupt storm. Chipsets may have problems based on boot, how the BIOS configures interrupts before correctness of the APIC (MADT) table, and routing of the System Control Interrupt (SCI). interrupt storms Interrupt storms can be distinguished from lost interrupts by checking the output of vmstat -i and looking at the line that has acpi0. If the counter is increasing at more than a couple per second, there is an interrupt storm. If the system appears hung, try breaking to DDB ( CTRL ALT ESC on console) and type show interrupts. APIC disabling When dealing with interrupt problems, try disabling APIC support with hint.apic.0.disabled="1" in /boot/loader.conf. Panics Panics are relatively rare for ACPI and are the top priority to be fixed. The first step is to isolate the steps to reproduce the panic, if possible, and get a backtrace. Follow the advice for enabling options DDB and setting up a serial console in or setting up a dump partition. To get a backtrace in DDB, use tr. When handwriting the backtrace, get at least the last five and the top five lines in the trace. Then, try to isolate the problem by booting with ACPI disabled. If that works, isolate the ACPI subsystem by using various values of . See &man.acpi.4; for some examples. System Powers Up After Suspend or Shutdown First, try setting hw.acpi.disable_on_poweroff="0" in /boot/loader.conf. This keeps ACPI from disabling various events during the shutdown process. Some systems need this value set to 1 (the default) for the same reason. This usually fixes the problem of a system powering up spontaneously after a suspend or poweroff. BIOS Contains Buggy Bytecode ACPI ASL Some BIOS vendors provide incorrect or buggy bytecode. This is usually manifested by kernel console messages like this: ACPI-1287: *** Error: Method execution failed [\\_SB_.PCI0.LPC0.FIGD._STA] \\ (Node 0xc3f6d160), AE_NOT_FOUND Often, these problems may be resolved by updating the BIOS to the latest revision. Most console messages are harmless, but if there are other problems, like the battery status is not working, these messages are a good place to start looking for problems. Overriding the Default <acronym>AML</acronym> The BIOS bytecode, known as ACPI Machine Language (AML), is compiled from a source language called ACPI Source Language (ASL). The AML is found in the table known as the Differentiated System Description Table (DSDT). ACPI ASL The goal of &os; is for everyone to have working ACPI without any user intervention. Workarounds are still being developed for common mistakes made by BIOS vendors. The µsoft; interpreter (acpi.sys and acpiec.sys) does not strictly check for adherence to the standard, and thus many BIOS vendors who only test ACPI under &windows; never fix their ASL. &os; developers continue to identify and document which non-standard behavior is allowed by µsoft;'s interpreter and replicate it so that &os; can work without forcing users to fix the ASL. To help identify buggy behavior and possibly fix it manually, a copy can be made of the system's ASL. To copy the system's ASL to a specified file name, use acpidump with , to show the contents of the fixed tables, and , to disassemble the AML: &prompt.root; acpidump -td > my.asl Some AML versions assume the user is running &windows;. To override this, set hw.acpi.osname="Windows 2009" in /boot/loader.conf, using the most recent &windows; version listed in the ASL. Other workarounds may require my.asl to be customized. If this file is edited, compile the new ASL using the following command. Warnings can usually be ignored, but errors are bugs that will usually prevent ACPI from working correctly. &prompt.root; iasl -f my.asl Including forces creation of the AML, even if there are errors during compilation. Some errors, such as missing return statements, are automatically worked around by the &os; interpreter. The default output filename for iasl is DSDT.aml. Load this file instead of the BIOS's buggy copy, which is still present in flash memory, by editing /boot/loader.conf as follows: acpi_dsdt_load="YES" acpi_dsdt_name="/boot/DSDT.aml" Be sure to copy DSDT.aml to /boot, then reboot the system. If this fixes the problem, send a &man.diff.1; of the old and new ASL to &a.acpi.name; so that developers can work around the buggy behavior in acpica. Getting and Submitting Debugging Info Nate Lawson Written by Peter Schultz With contributions from Tom Rhodes ACPI problems ACPI debugging The ACPI driver has a flexible debugging facility. A set of subsystems and the level of verbosity can be specified. The subsystems to debug are specified as layers and are broken down into components (ACPI_ALL_COMPONENTS) and ACPI hardware support (ACPI_ALL_DRIVERS). The verbosity of debugging output is specified as the level and ranges from just report errors (ACPI_LV_ERROR) to everything (ACPI_LV_VERBOSE). The level is a bitmask so multiple options can be set at once, separated by spaces. In practice, a serial console should be used to log the output so it is not lost as the console message buffer flushes. A full list of the individual layers and levels is found in &man.acpi.4;. Debugging output is not enabled by default. To enable it, add options ACPI_DEBUG to the custom kernel configuration file if ACPI is compiled into the kernel. Add ACPI_DEBUG=1 to /etc/make.conf to enable it globally. If a module is used instead of a custom kernel, recompile just the acpi.ko module as follows: &prompt.root; cd /sys/modules/acpi/acpi && make clean && make ACPI_DEBUG=1 Copy the compiled acpi.ko to /boot/kernel and add the desired level and layer to /boot/loader.conf. The entries in this example enable debug messages for all ACPI components and hardware drivers and output error messages at the least verbose level: debug.acpi.layer="ACPI_ALL_COMPONENTS ACPI_ALL_DRIVERS" debug.acpi.level="ACPI_LV_ERROR" If the required information is triggered by a specific event, such as a suspend and then resume, do not modify /boot/loader.conf. Instead, use sysctl to specify the layer and level after booting and preparing the system for the specific event. The variables which can be set using sysctl are named the same as the tunables in /boot/loader.conf. ACPI problems Once the debugging information is gathered, it can be sent to &a.acpi.name; so that it can be used by the &os; ACPI maintainers to identify the root cause of the problem and to develop a solution. Before submitting debugging information to this mailing list, ensure the latest BIOS version is installed and, if available, the embedded controller firmware version. When submitting a problem report, include the following information: Description of the buggy behavior, including system type, model, and anything that causes the bug to appear. Note as accurately as possible when the bug began occurring if it is new. The output of dmesg after running boot -v, including any error messages generated by the bug. The dmesg output from boot -v with ACPI disabled, if disabling ACPI helps to fix the problem. Output from sysctl hw.acpi. This lists which features the system offers. The URL to a pasted version of the system's ASL. Do not send the ASL directly to the list as it can be very large. Generate a copy of the ASL by running this command: &prompt.root; acpidump -dt > name-system.asl Substitute the login name for name and manufacturer/model for system. For example, use njl-FooCo6000.asl. Most &os; developers watch the &a.current;, but one should submit problems to &a.acpi.name; to be sure it is seen. Be patient when waiting for a response. If the bug is not immediately apparent, submit a PR using &man.send-pr.1;. When entering a PR, include the same information as requested above. This helps developers to track the problem and resolve it. Do not send a PR without emailing &a.acpi.name; first as it is likely that the problem has been reported before. References More information about ACPI may be found in the following locations: The &os; ACPI Mailing List Archives (http://lists.freebsd.org/pipermail/freebsd-acpi/) The ACPI 2.0 Specification (http://acpi.info/spec.htm) &man.acpi.4;, &man.acpi.thermal.4;, &man.acpidump.8;, &man.iasl.8;, and &man.acpidb.8; Index: head/en_US.ISO8859-1/books/handbook/disks/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/disks/chapter.xml (revision 50807) +++ head/en_US.ISO8859-1/books/handbook/disks/chapter.xml (revision 50808) @@ -1,3806 +1,3806 @@ Storage Synopsis This chapter covers the use of disks and storage media in &os;. This includes SCSI and IDE disks, CD and DVD media, memory-backed disks, and USB storage devices. After reading this chapter, you will know: How to add additional hard disks to a &os; system. How to grow the size of a disk's partition on &os;. How to configure &os; to use USB storage devices. How to use CD and DVD media on a &os; system. How to use the backup programs available under &os;. How to set up memory disks. What file system snapshots are and how to use them efficiently. How to use quotas to limit disk space usage. How to encrypt disks and swap to secure them against attackers. How to configure a highly available storage network. Before reading this chapter, you should: Know how to configure and install a new &os; kernel. Adding Disks David O'Brien Originally contributed by disks adding This section describes how to add a new SATA disk to a machine that currently only has a single drive. First, turn off the computer and install the drive in the computer following the instructions of the computer, controller, and drive manufacturers. Reboot the system and become root. Inspect /var/run/dmesg.boot to ensure the new disk was found. In this example, the newly added SATA drive will appear as ada1. partitions gpart For this example, a single large partition will be created on the new disk. The GPT partitioning scheme will be used in preference to the older and less versatile MBR scheme. If the disk to be added is not blank, old partition information can be removed with gpart delete. See &man.gpart.8; for details. The partition scheme is created, and then a single partition is added. To improve performance on newer disks with larger hardware block sizes, the partition is aligned to one megabyte boundaries: &prompt.root; gpart create -s GPT ada1 &prompt.root; gpart add -t freebsd-ufs -a 1M ada1 Depending on use, several smaller partitions may be desired. See &man.gpart.8; for options to create partitions smaller than a whole disk. The disk partition information can be viewed with gpart show: &prompt.user; gpart show ada1 => 34 1465146988 ada1 GPT (699G) 34 2014 - free - (1.0M) 2048 1465143296 1 freebsd-ufs (699G) 1465145344 1678 - free - (839K) A file system is created in the new partition on the new disk: &prompt.root; newfs -U /dev/ada1p1 An empty directory is created as a mountpoint, a location for mounting the new disk in the original disk's file system: &prompt.root; mkdir /newdisk Finally, an entry is added to /etc/fstab so the new disk will be mounted automatically at startup: /dev/ada1p1 /newdisk ufs rw 2 2 The new disk can be mounted manually, without restarting the system: &prompt.root; mount /newdisk Resizing and Growing Disks Allan Jude Originally contributed by disks resizing A disk's capacity can increase without any changes to the data already present. This happens commonly with virtual machines, when the virtual disk turns out to be too small and is enlarged. Sometimes a disk image is written to a USB memory stick, but does not use the full capacity. Here we describe how to resize or grow disk contents to take advantage of increased capacity. Determine the device name of the disk to be resized by inspecting /var/run/dmesg.boot. In this example, there is only one SATA disk in the system, so the drive will appear as ada0. partitions gpart List the partitions on the disk to see the current configuration: &prompt.root; gpart show ada0 => 34 83886013 ada0 GPT (48G) [CORRUPT] 34 128 1 freebsd-boot (64k) 162 79691648 2 freebsd-ufs (38G) 79691810 4194236 3 freebsd-swap (2G) 83886046 1 - free - (512B) If the disk was formatted with the GPT partitioning scheme, it may show as corrupted because the GPT backup partition table is no longer at the end of the drive. Fix the backup partition table with gpart: &prompt.root; gpart recover ada0 ada0 recovered Now the additional space on the disk is available for use by a new partition, or an existing partition can be expanded: &prompt.root; gpart show ada0 => 34 102399933 ada0 GPT (48G) 34 128 1 freebsd-boot (64k) 162 79691648 2 freebsd-ufs (38G) 79691810 4194236 3 freebsd-swap (2G) 83886046 18513921 - free - (8.8G) Partitions can only be resized into contiguous free space. Here, the last partition on the disk is the swap partition, but the second partition is the one that needs to be resized. Swap partitions only contain temporary data, so it can safely be unmounted, deleted, and then recreated after resizing other partitions. &prompt.root; swapoff /dev/ada0p3 &prompt.root; gpart delete -i 3 ada0 ada0p3 deleted &prompt.root; gpart show ada0 => 34 102399933 ada0 GPT (48G) 34 128 1 freebsd-boot (64k) 162 79691648 2 freebsd-ufs (38G) 79691810 22708157 - free - (10G) There is risk of data loss when modifying the partition table of a mounted file system. It is best to perform the following steps on an unmounted file system while running from a live CD-ROM or USB device. However, if absolutely necessary, a mounted file system can be resized after disabling GEOM safety features: &prompt.root; sysctl kern.geom.debugflags=16 Resize the partition, leaving room to recreate a swap partition of the desired size. This only modifies the size of the partition. The file system in the partition will be expanded in a separate step. &prompt.root; gpart resize -i 2 -a 4k -s 47G ada0 ada0p2 resized &prompt.root; gpart show ada0 => 34 102399933 ada0 GPT (48G) 34 128 1 freebsd-boot (64k) 162 98566144 2 freebsd-ufs (47G) 98566306 3833661 - free - (1.8G) Recreate the swap partition: &prompt.root; gpart add -t freebsd-swap -a 4k ada0 ada0p3 added &prompt.root; gpart show ada0 => 34 102399933 ada0 GPT (48G) 34 128 1 freebsd-boot (64k) 162 98566144 2 freebsd-ufs (47G) 98566306 3833661 3 freebsd-swap (1.8G) &prompt.root; swapon /dev/ada0p3 Grow the UFS file system to use the new capacity of the resized partition: Growing a live UFS file system is only possible in &os; 10.0-RELEASE and later. For earlier versions, the file system must not be mounted. &prompt.root; growfs /dev/ada0p2 Device is mounted read-write; resizing will result in temporary write suspension for /. It's strongly recommended to make a backup before growing the file system. OK to grow file system on /dev/ada0p2, mounted on /, from 38GB to 47GB? [Yes/No] Yes super-block backups (for fsck -b #) at: 80781312, 82063552, 83345792, 84628032, 85910272, 87192512, 88474752, 89756992, 91039232, 92321472, 93603712, 94885952, 96168192, 97450432 Both the partition and the file system on it have now been resized to use the newly-available disk space. <acronym>USB</acronym> Storage Devices Marc Fonvieille Contributed by USB disks Many external storage solutions, such as hard drives, USB thumbdrives, and CD and DVD burners, use the Universal Serial Bus (USB). &os; provides support for USB 1.x, 2.0, and 3.0 devices. USB 3.0 support is not compatible with some hardware, including Haswell (Lynx point) chipsets. If &os; boots with a failed with error 19 message, disable xHCI/USB3 in the system BIOS. Support for USB storage devices is built into the GENERIC kernel. For a custom kernel, be sure that the following lines are present in the kernel configuration file: device scbus # SCSI bus (required for ATA/SCSI) device da # Direct Access (disks) device pass # Passthrough device (direct ATA/SCSI access) device uhci # provides USB 1.x support device ohci # provides USB 1.x support device ehci # provides USB 2.0 support device xhci # provides USB 3.0 support device usb # USB Bus (required) device umass # Disks/Mass storage - Requires scbus and da device cd # needed for CD and DVD burners &os; uses the &man.umass.4; driver which uses the SCSI subsystem to access USB storage devices. Since any USB device will be seen as a SCSI device by the system, if the USB device is a CD or DVD burner, do not include in a custom kernel configuration file. The rest of this section demonstrates how to verify that a USB storage device is recognized by &os; and how to configure the device so that it can be used. Device Configuration To test the USB configuration, plug in the USB device. Use dmesg to confirm that the drive appears in the system message buffer. It should look something like this: umass0: <STECH Simple Drive, class 0/0, rev 2.00/1.04, addr 3> on usbus0 umass0: SCSI over Bulk-Only; quirks = 0x0100 umass0:4:0:-1: Attached to scbus4 da0 at umass-sim0 bus 0 scbus4 target 0 lun 0 da0: <STECH Simple Drive 1.04> Fixed Direct Access SCSI-4 device da0: Serial Number WD-WXE508CAN263 da0: 40.000MB/s transfers da0: 152627MB (312581808 512 byte sectors: 255H 63S/T 19457C) da0: quirks=0x2<NO_6_BYTE> The brand, device node (da0), speed, and size will differ according to the device. Since the USB device is seen as a SCSI one, camcontrol can be used to list the USB storage devices attached to the system: &prompt.root; camcontrol devlist <STECH Simple Drive 1.04> at scbus4 target 0 lun 0 (pass3,da0) Alternately, usbconfig can be used to list the device. Refer to &man.usbconfig.8; for more information about this command. &prompt.root; usbconfig ugen0.3: <Simple Drive STECH> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps) pwr=ON (2mA) If the device has not been formatted, refer to for instructions on how to format and create partitions on the USB drive. If the drive comes with a file system, it can be mounted by root using the instructions in . Allowing untrusted users to mount arbitrary media, by enabling vfs.usermount as described below, should not be considered safe from a security point of view. Most file systems were not built to safeguard against malicious devices. To make the device mountable as a normal user, one solution is to make all users of the device a member of the operator group using &man.pw.8;. Next, ensure that operator is able to read and write the device by adding these lines to /etc/devfs.rules: [localrules=5] add path 'da*' mode 0660 group operator If internal SCSI disks are also installed in the system, change the second line as follows: add path 'da[3-9]*' mode 0660 group operator This will exclude the first three SCSI disks (da0 to da2)from belonging to the operator group. Replace 3 with the number of internal SCSI disks. Refer to &man.devfs.rules.5; for more information about this file. Next, enable the ruleset in /etc/rc.conf: devfs_system_ruleset="localrules" Then, instruct the system to allow regular users to mount file systems by adding the following line to /etc/sysctl.conf: vfs.usermount=1 Since this only takes effect after the next reboot, use sysctl to set this variable now: &prompt.root; sysctl vfs.usermount=1 vfs.usermount: 0 -> 1 The final step is to create a directory where the file system is to be mounted. This directory needs to be owned by the user that is to mount the file system. One way to do that is for root to create a subdirectory owned by that user as /mnt/username. In the following example, replace username with the login name of the user and usergroup with the user's primary group: &prompt.root; mkdir /mnt/username &prompt.root; chown username:usergroup /mnt/username Suppose a USB thumbdrive is plugged in, and a device /dev/da0s1 appears. If the device is formatted with a FAT file system, the user can mount it using: &prompt.user; mount -t msdosfs -o -m=644,-M=755 /dev/da0s1 /mnt/username Before the device can be unplugged, it must be unmounted first: &prompt.user; umount /mnt/username After device removal, the system message buffer will show messages similar to the following: umass0: at uhub3, port 2, addr 3 (disconnected) da0 at umass-sim0 bus 0 scbus4 target 0 lun 0 da0: <STECH Simple Drive 1.04> s/n WD-WXE508CAN263 detached (da0:umass-sim0:0:0:0): Periph destroyed Automounting Removable Media &man.autofs.5; supports automatic mounting of removable media starting with &os; 10.2-RELEASE. USB devices can be automatically mounted by uncommenting this line in /etc/auto_master: /media -media -nosuid Then add these lines to /etc/devd.conf: notify 100 { match "system" "GEOM"; match "subsystem" "DEV"; action "/usr/sbin/automount -c"; }; Reload the configuration if &man.autofs.5; and &man.devd.8; are already running: &prompt.root; service automount reload &prompt.root; service devd restart &man.autofs.5; can be set to start at boot by adding this line to /etc/rc.conf: autofs_enable="YES" &man.autofs.5; requires &man.devd.8; to be enabled, as it is by default. Start the services immediately with: &prompt.root; service automount start &prompt.root; service automountd start &prompt.root; service autounmountd start &prompt.root; service devd start Each file system that can be automatically mounted appears as a directory in /media/. The directory is named after the file system label. If the label is missing, the directory is named after the device node. The file system is transparently mounted on the first access, and unmounted after a period of inactivity. Automounted drives can also be unmounted manually: &prompt.root; automount -fu This mechanism is typically used for memory cards and USB memory sticks. It can be used with any block device, including optical drives or iSCSI LUNs. <acronym>USB</acronym> Mass Storage Target The &man.cfumass.4; driver is a USB device mode driver first available in &os; 12.0. When running on USB OTG-compliant hardware like that built into many embedded boards, the &os; USB stack can run in device mode. Device mode makes it possible for the computer to present itself as different kinds of USB device classes, including serial ports, network adapters, and mass storage. A USB host like a laptop or desktop computer is able to access them just like physical USB devices. The &man.usb.template.4; kernel module allows the USB stack to switch between host-side and device-side automatically, depending on what is connected to the USB port. Connecting a USB device like a memory stick to the USB OTG port causes &os; to switch to host mode. Connecting a USB host like a computer causes &os; to switch to device mode. What &os; presents to the USB host depends on the hw.usb.template sysctl. See &man.usb.template.4; for the list of available values. Note that for the host to notice the configuration change, it must be either physically disconnected and reconnected, or forced to rescan the USB bus in a system-specific way. When &os; is running on the host, &man.usbconfig.8; reset can be used. This also must be done after loading usb_template.ko if the USB host was already connected to the USB OTG socket. The hw.usb.template sysctl is set to 0 by default, making &os; work as a USB Mass Storage target. Both &man.usb.template.4; and &man.cfumass.4; kernel modules must be loaded. &man.cfumass.4; interfaces to the CTL subsystem, the same one that is used for iSCSI or Fibre Channel targets. On the host side, USB Mass Storage initiators can only access a single LUN, LUN 0. USB Mass Storage does not require the &man.ctld.8; daemon to be running, although it can be used if desired. This is different from iSCSI. Thus, there are two ways to configure the target: &man.ctladm.8;, or &man.ctld.8;. Both require the cfumass.ko kernel module to be loaded. The module can be loaded manually: &prompt.root; kldload cfumass If cfumass.ko has not been built into the kernel, /boot/loader.conf can be set to load the module at boot: cfumass_load="YES" A LUN can be created without the &man.ctld.8; daemon: &prompt.root; ctladm create -b block -o file=/data/target0 This presents the contents of the image file /data/target0 as a LUN to the USB host. The file must exist before executing the command. To configure the LUN at system startup, add the command to /etc/rc.local. &man.ctld.8; can also be used to manage LUNs. Create /etc/ctl.conf, add a line to /etc/rc.conf to make sure &man.ctld.8; is automatically started at boot, and then start the daemon. This is an example of a simple /etc/ctl.conf configuration file. Refer to &man.ctl.conf.5; for a more complete description of the options. target naa.50015178f369f092 { lun 0 { path /data/target0 size 4G } } The example creates a single target with a single LUN. The naa.50015178f369f092 is a device identifier composed of 32 random hexadecimal digits. The path line defines the full path to a file or zvol backing the LUN. That file must exist before starting &man.ctld.8;. The second line is optional and specifies the size of the LUN. To make sure the &man.ctld.8; daemon is started at boot, add this line to /etc/rc.conf: ctld_enable="YES" To start &man.ctld.8; now, run this command: &prompt.root; service ctld start As the &man.ctld.8; daemon is started, it reads /etc/ctl.conf. If this file is edited after the daemon starts, reload the changes so they take effect immediately: &prompt.root; service ctld reload Creating and Using <acronym>CD</acronym> Media Mike Meyer Contributed by CD-ROMs creating Compact Disc (CD) media provide a number of features that differentiate them from conventional disks. They are designed so that they can be read continuously without delays to move the head between tracks. While CD media do have tracks, these refer to a section of data to be read continuously, and not a physical property of the disk. The ISO 9660 file system was designed to deal with these differences. ISO 9660 file systems ISO 9660 CD burner ATAPI The &os; Ports Collection provides several utilities for burning and duplicating audio and data CDs. This chapter demonstrates the use of several command line utilities. For CD burning software with a graphical utility, consider installing the sysutils/xcdroast or sysutils/k3b packages or ports. Supported Devices Marc Fonvieille Contributed by CD burner ATAPI/CAM driver The GENERIC kernel provides support for SCSI, USB, and ATAPI CD readers and burners. If a custom kernel is used, the options that need to be present in the kernel configuration file vary by the type of device. For a SCSI burner, make sure these options are present: device scbus # SCSI bus (required for ATA/SCSI) device da # Direct Access (disks) device pass # Passthrough device (direct ATA/SCSI access) device cd # needed for CD and DVD burners For a USB burner, make sure these options are present: device scbus # SCSI bus (required for ATA/SCSI) device da # Direct Access (disks) device pass # Passthrough device (direct ATA/SCSI access) device cd # needed for CD and DVD burners device uhci # provides USB 1.x support device ohci # provides USB 1.x support device ehci # provides USB 2.0 support device xhci # provides USB 3.0 support device usb # USB Bus (required) device umass # Disks/Mass storage - Requires scbus and da For an ATAPI burner, make sure these options are present: device ata # Legacy ATA/SATA controllers device scbus # SCSI bus (required for ATA/SCSI) device pass # Passthrough device (direct ATA/SCSI access) device cd # needed for CD and DVD burners On &os; versions prior to 10.x, this line is also needed in the kernel configuration file if the burner is an ATAPI device: device atapicam Alternately, this driver can be loaded at boot time by adding the following line to /boot/loader.conf: atapicam_load="YES" This will require a reboot of the system as this driver can only be loaded at boot time. To verify that &os; recognizes the device, run dmesg and look for an entry for the device. On systems prior to 10.x, the device name in the first line of the output will be acd0 instead of cd0. &prompt.user; dmesg | grep cd cd0 at ahcich1 bus 0 scbus1 target 0 lun 0 cd0: <HL-DT-ST DVDRAM GU70N LT20> Removable CD-ROM SCSI-0 device cd0: Serial Number M3OD3S34152 cd0: 150.000MB/s transfers (SATA 1.x, UDMA6, ATAPI 12bytes, PIO 8192bytes) cd0: Attempt to query device size failed: NOT READY, Medium not present - tray closed Burning a <acronym>CD</acronym> In &os;, cdrecord can be used to burn CDs. This command is installed with the sysutils/cdrtools package or port. While cdrecord has many options, basic usage is simple. Specify the name of the ISO file to burn and, if the system has multiple burner devices, specify the name of the device to use: &prompt.root; cdrecord dev=device imagefile.iso To determine the device name of the burner, use which might produce results like this: CD-ROMs burning &prompt.root; cdrecord -scanbus ProDVD-ProBD-Clone 3.00 (amd64-unknown-freebsd10.0) Copyright (C) 1995-2010 Jörg Schilling Using libscg version 'schily-0.9' scsibus0: 0,0,0 0) 'SEAGATE ' 'ST39236LW ' '0004' Disk 0,1,0 1) 'SEAGATE ' 'ST39173W ' '5958' Disk 0,2,0 2) * 0,3,0 3) 'iomega ' 'jaz 1GB ' 'J.86' Removable Disk 0,4,0 4) 'NEC ' 'CD-ROM DRIVE:466' '1.26' Removable CD-ROM 0,5,0 5) * 0,6,0 6) * 0,7,0 7) * scsibus1: 1,0,0 100) * 1,1,0 101) * 1,2,0 102) * 1,3,0 103) * 1,4,0 104) * 1,5,0 105) 'YAMAHA ' 'CRW4260 ' '1.0q' Removable CD-ROM 1,6,0 106) 'ARTEC ' 'AM12S ' '1.06' Scanner 1,7,0 107) * Locate the entry for the CD burner and use the three numbers separated by commas as the value for . In this case, the Yamaha burner device is 1,5,0, so the appropriate input to specify that device is . Refer to the manual page for cdrecord for other ways to specify this value and for information on writing audio tracks and controlling the write speed. Alternately, run the following command to get the device address of the burner: &prompt.root; camcontrol devlist <MATSHITA CDRW/DVD UJDA740 1.00> at scbus1 target 0 lun 0 (cd0,pass0) Use the numeric values for scbus, target, and lun. For this example, 1,0,0 is the device name to use. Writing Data to an <acronym>ISO</acronym> File System In order to produce a data CD, the data files that are going to make up the tracks on the CD must be prepared before they can be burned to the CD. In &os;, sysutils/cdrtools installs mkisofs, which can be used to produce an ISO 9660 file system that is an image of a directory tree within a &unix; file system. The simplest usage is to specify the name of the ISO file to create and the path to the files to place into the ISO 9660 file system: &prompt.root; mkisofs -o imagefile.iso /path/to/tree file systems ISO 9660 This command maps the file names in the specified path to names that fit the limitations of the standard ISO 9660 file system, and will exclude files that do not meet the standard for ISO file systems. file systems Joliet A number of options are available to overcome the restrictions imposed by the standard. In particular, enables the Rock Ridge extensions common to &unix; systems and enables Joliet extensions used by µsoft; systems. For CDs that are going to be used only on &os; systems, can be used to disable all filename restrictions. When used with , it produces a file system image that is identical to the specified &os; tree, even if it violates the ISO 9660 standard. CD-ROMs creating bootable The last option of general use is . This is used to specify the location of a boot image for use in producing an El Torito bootable CD. This option takes an argument which is the path to a boot image from the top of the tree being written to the CD. By default, mkisofs creates an ISO image in floppy disk emulation mode, and thus expects the boot image to be exactly 1200, 1440 or 2880 KB in size. Some boot loaders, like the one used by the &os; distribution media, do not use emulation mode. In this case, should be used. So, if /tmp/myboot holds a bootable &os; system with the boot image in /tmp/myboot/boot/cdboot, this command would produce /tmp/bootable.iso: &prompt.root; mkisofs -R -no-emul-boot -b boot/cdboot -o /tmp/bootable.iso /tmp/myboot The resulting ISO image can be mounted as a memory disk with: &prompt.root; mdconfig -a -t vnode -f /tmp/bootable.iso -u 0 &prompt.root; mount -t cd9660 /dev/md0 /mnt One can then verify that /mnt and /tmp/myboot are identical. There are many other options available for mkisofs to fine-tune its behavior. Refer to &man.mkisofs.8; for details. It is possible to copy a data CD to an image file that is functionally equivalent to the image file created with mkisofs. To do so, use dd with the device name as the input file and the name of the ISO to create as the output file: &prompt.root; dd if=/dev/cd0 of=file.iso bs=2048 The resulting image file can be burned to CD as described in . Using Data <acronym>CD</acronym>s Once an ISO has been burned to a CD, it can be mounted by specifying the file system type, the name of the device containing the CD, and an existing mount point: &prompt.root; mount -t cd9660 /dev/cd0 /mnt Since mount assumes that a file system is of type ufs, a Incorrect super block error will occur if -t cd9660 is not included when mounting a data CD. While any data CD can be mounted this way, disks with certain ISO 9660 extensions might behave oddly. For example, Joliet disks store all filenames in two-byte Unicode characters. If some non-English characters show up as question marks, specify the local charset with . For more information, refer to &man.mount.cd9660.8;. In order to do this character conversion with the help of , the kernel requires the cd9660_iconv.ko module to be loaded. This can be done either by adding this line to loader.conf: cd9660_iconv_load="YES" and then rebooting the machine, or by directly loading the module with kldload. Occasionally, Device not configured will be displayed when trying to mount a data CD. This usually means that the CD drive has not detected a disk in the tray, or that the drive is not visible on the bus. It can take a couple of seconds for a CD drive to detect media, so be patient. Sometimes, a SCSI CD drive may be missed because it did not have enough time to answer the bus reset. To resolve this, a custom kernel can be created which increases the default SCSI delay. Add the following option to the custom kernel configuration file and rebuild the kernel using the instructions in : options SCSI_DELAY=15000 This tells the SCSI bus to pause 15 seconds during boot, to give the CD drive every possible chance to answer the bus reset. It is possible to burn a file directly to CD, without creating an ISO 9660 file system. This is known as burning a raw data CD and some people do this for backup purposes. This type of disk can not be mounted as a normal data CD. In order to retrieve the data burned to such a CD, the data must be read from the raw device node. For example, this command will extract a compressed tar file located on the second CD device into the current working directory: &prompt.root; tar xzvf /dev/cd1 In order to mount a data CD, the data must be written using mkisofs. Duplicating Audio <acronym>CD</acronym>s To duplicate an audio CD, extract the audio data from the CD to a series of files, then write these files to a blank CD. describes how to duplicate and burn an audio CD. If the &os; version is less than 10.0 and the device is ATAPI, the module must be first loaded using the instructions in . Duplicating an Audio <acronym>CD</acronym> The sysutils/cdrtools package or port installs cdda2wav. This command can be used to extract all of the audio tracks, with each track written to a separate WAV file in the current working directory: &prompt.user; cdda2wav -vall -B -Owav A device name does not need to be specified if there is only one CD device on the system. Refer to the cdda2wav manual page for instructions on how to specify a device and to learn more about the other options available for this command. Use cdrecord to write the .wav files: &prompt.user; cdrecord -v dev=2,0 -dao -useinfo *.wav Make sure that 2,0 is set appropriately, as described in . Creating and Using <acronym>DVD</acronym> Media Marc Fonvieille Contributed by Andy Polyakov With inputs from DVD burning Compared to the CD, the DVD is the next generation of optical media storage technology. The DVD can hold more data than any CD and is the standard for video publishing. Five physical recordable formats can be defined for a recordable DVD: DVD-R: This was the first DVD recordable format available. The DVD-R standard is defined by the DVD Forum. This format is write once. DVD-RW: This is the rewritable version of the DVD-R standard. A DVD-RW can be rewritten about 1000 times. DVD-RAM: This is a rewritable format which can be seen as a removable hard drive. However, this media is not compatible with most DVD-ROM drives and DVD-Video players as only a few DVD writers support the DVD-RAM format. Refer to for more information on DVD-RAM use. DVD+RW: This is a rewritable format defined by the DVD+RW Alliance. A DVD+RW can be rewritten about 1000 times. DVD+R: This format is the write once variation of the DVD+RW format. A single layer recordable DVD can hold up to 4,700,000,000 bytes which is actually 4.38 GB or 4485 MB as 1 kilobyte is 1024 bytes. A distinction must be made between the physical media and the application. For example, a DVD-Video is a specific file layout that can be written on any recordable DVD physical media such as DVD-R, DVD+R, or DVD-RW. Before choosing the type of media, ensure that both the burner and the DVD-Video player are compatible with the media under consideration. Configuration To perform DVD recording, use &man.growisofs.1;. This command is part of the sysutils/dvd+rw-tools utilities which support all DVD media types. These tools use the SCSI subsystem to access the devices, therefore ATAPI/CAM support must be loaded or statically compiled into the kernel. This support is not needed if the burner uses the USB interface. Refer to for more details on USB device configuration. DMA access must also be enabled for ATAPI devices, by adding the following line to /boot/loader.conf: hw.ata.atapi_dma="1" Before attempting to use dvd+rw-tools, consult the Hardware Compatibility Notes. For a graphical user interface, consider using sysutils/k3b which provides a user friendly interface to &man.growisofs.1; and many other burning tools. Burning Data <acronym>DVD</acronym>s Since &man.growisofs.1; is a front-end to mkisofs, it will invoke &man.mkisofs.8; to create the file system layout and perform the write on the DVD. This means that an image of the data does not need to be created before the burning process. To burn to a DVD+R or a DVD-R the data in /path/to/data, use the following command: &prompt.root; growisofs -dvd-compat -Z /dev/cd0 -J -R /path/to/data In this example, is passed to &man.mkisofs.8; to create an ISO 9660 file system with Joliet and Rock Ridge extensions. Refer to &man.mkisofs.8; for more details. For the initial session recording, is used for both single and multiple sessions. Replace /dev/cd0, with the name of the DVD device. Using indicates that the disk will be closed and that the recording will be unappendable. This should also provide better media compatibility with DVD-ROM drives. To burn a pre-mastered image, such as imagefile.iso, use: &prompt.root; growisofs -dvd-compat -Z /dev/cd0=imagefile.iso The write speed should be detected and automatically set according to the media and the drive being used. To force the write speed, use . Refer to &man.growisofs.1; for example usage. In order to support working files larger than 4.38GB, an UDF/ISO-9660 hybrid file system must be created by passing to &man.mkisofs.8; and all related programs, such as &man.growisofs.1;. This is required only when creating an ISO image file or when writing files directly to a disk. Since a disk created this way must be mounted as an UDF file system with &man.mount.udf.8;, it will be usable only on an UDF aware operating system. Otherwise it will look as if it contains corrupted files. To create this type of ISO file: &prompt.user; mkisofs -R -J -udf -iso-level 3 -o imagefile.iso /path/to/data To burn files directly to a disk: &prompt.root; growisofs -dvd-compat -udf -iso-level 3 -Z /dev/cd0 -J -R /path/to/data When an ISO image already contains large files, no additional options are required for &man.growisofs.1; to burn that image on a disk. Be sure to use an up-to-date version of sysutils/cdrtools, which contains &man.mkisofs.8;, as an older version may not contain large files support. If the latest version does not work, install sysutils/cdrtools-devel and read its &man.mkisofs.8;. Burning a <acronym>DVD</acronym>-Video DVD DVD-Video A DVD-Video is a specific file layout based on the ISO 9660 and micro-UDF (M-UDF) specifications. Since DVD-Video presents a specific data structure hierarchy, a particular program such as multimedia/dvdauthor is needed to author the DVD. If an image of the DVD-Video file system already exists, it can be burned in the same way as any other image. If dvdauthor was used to make the DVD and the result is in /path/to/video, the following command should be used to burn the DVD-Video: &prompt.root; growisofs -Z /dev/cd0 -dvd-video /path/to/video is passed to &man.mkisofs.8; to instruct it to create a DVD-Video file system layout. This option implies the &man.growisofs.1; option. Using a <acronym>DVD+RW</acronym> DVD DVD+RW Unlike CD-RW, a virgin DVD+RW needs to be formatted before first use. It is recommended to let &man.growisofs.1; take care of this automatically whenever appropriate. However, it is possible to use dvd+rw-format to format the DVD+RW: &prompt.root; dvd+rw-format /dev/cd0 Only perform this operation once and keep in mind that only virgin DVD+RW medias need to be formatted. Once formatted, the DVD+RW can be burned as usual. To burn a totally new file system and not just append some data onto a DVD+RW, the media does not need to be blanked first. Instead, write over the previous recording like this: &prompt.root; growisofs -Z /dev/cd0 -J -R /path/to/newdata The DVD+RW format supports appending data to a previous recording. This operation consists of merging a new session to the existing one as it is not considered to be multi-session writing. &man.growisofs.1; will grow the ISO 9660 file system present on the media. For example, to append data to a DVD+RW, use the following: &prompt.root; growisofs -M /dev/cd0 -J -R /path/to/nextdata The same &man.mkisofs.8; options used to burn the initial session should be used during next writes. Use for better media compatibility with DVD-ROM drives. When using DVD+RW, this option will not prevent the addition of data. To blank the media, use: &prompt.root; growisofs -Z /dev/cd0=/dev/zero Using a <acronym>DVD-RW</acronym> DVD DVD-RW A DVD-RW accepts two disc formats: incremental sequential and restricted overwrite. By default, DVD-RW discs are in sequential format. A virgin DVD-RW can be directly written without being formatted. However, a non-virgin DVD-RW in sequential format needs to be blanked before writing a new initial session. To blank a DVD-RW in sequential mode: &prompt.root; dvd+rw-format -blank=full /dev/cd0 A full blanking using will take about one hour on a 1x media. A fast blanking can be performed using , if the DVD-RW will be recorded in Disk-At-Once (DAO) mode. To burn the DVD-RW in DAO mode, use the command: &prompt.root; growisofs -use-the-force-luke=dao -Z /dev/cd0=imagefile.iso Since &man.growisofs.1; automatically attempts to detect fast blanked media and engage DAO write, should not be required. One should instead use restricted overwrite mode with any DVD-RW as this format is more flexible than the default of incremental sequential. To write data on a sequential DVD-RW, use the same instructions as for the other DVD formats: &prompt.root; growisofs -Z /dev/cd0 -J -R /path/to/data To append some data to a previous recording, use with &man.growisofs.1;. However, if data is appended on a DVD-RW in incremental sequential mode, a new session will be created on the disc and the result will be a multi-session disc. A DVD-RW in restricted overwrite format does not need to be blanked before a new initial session. Instead, overwrite the disc with . It is also possible to grow an existing ISO 9660 file system written on the disc with . The result will be a one-session DVD. To put a DVD-RW in restricted overwrite format, the following command must be used: &prompt.root; dvd+rw-format /dev/cd0 To change back to sequential format, use: &prompt.root; dvd+rw-format -blank=full /dev/cd0 Multi-Session Few DVD-ROM drives support multi-session DVDs and most of the time only read the first session. DVD+R, DVD-R and DVD-RW in sequential format can accept multiple sessions. The notion of multiple sessions does not exist for the DVD+RW and the DVD-RW restricted overwrite formats. Using the following command after an initial non-closed session on a DVD+R, DVD-R, or DVD-RW in sequential format, will add a new session to the disc: &prompt.root; growisofs -M /dev/cd0 -J -R /path/to/nextdata Using this command with a DVD+RW or a DVD-RW in restricted overwrite mode will append data while merging the new session to the existing one. The result will be a single-session disc. Use this method to add data after an initial write on these types of media. Since some space on the media is used between each session to mark the end and start of sessions, one should add sessions with a large amount of data to optimize media space. The number of sessions is limited to 154 for a DVD+R, about 2000 for a DVD-R, and 127 for a DVD+R Double Layer. For More Information To obtain more information about a DVD, use dvd+rw-mediainfo /dev/cd0 while the disc in the specified drive. More information about dvd+rw-tools can be found in &man.growisofs.1;, on the dvd+rw-tools web site, and in the cdwrite mailing list archives. When creating a problem report related to the use of dvd+rw-tools, always include the output of dvd+rw-mediainfo. Using a <acronym>DVD-RAM</acronym> DVD DVD-RAM DVD-RAM writers can use either a SCSI or ATAPI interface. For ATAPI devices, DMA access has to be enabled by adding the following line to /boot/loader.conf: hw.ata.atapi_dma="1" A DVD-RAM can be seen as a removable hard drive. Like any other hard drive, the DVD-RAM must be formatted before it can be used. In this example, the whole disk space will be formatted with a standard UFS2 file system: &prompt.root; dd if=/dev/zero of=/dev/acd0 bs=2k count=1 &prompt.root; bsdlabel -Bw acd0 &prompt.root; newfs /dev/acd0 The DVD device, acd0, must be changed according to the configuration. Once the DVD-RAM has been formatted, it can be mounted as a normal hard drive: &prompt.root; mount /dev/acd0 /mnt Once mounted, the DVD-RAM will be both readable and writeable. Creating and Using Floppy Disks This section explains how to format a 3.5 inch floppy disk in &os;. Steps to Format a Floppy A floppy disk needs to be low-level formatted before it can be used. This is usually done by the vendor, but formatting is a good way to check media integrity. To low-level format the floppy disk on &os;, use &man.fdformat.1;. When using this utility, make note of any error messages, as these can help determine if the disk is good or bad. To format the floppy, insert a new 3.5 inch floppy disk into the first floppy drive and issue: &prompt.root; /usr/sbin/fdformat -f 1440 /dev/fd0 After low-level formatting the disk, create a disk label as it is needed by the system to determine the size of the disk and its geometry. The supported geometry values are listed in /etc/disktab. To write the disk label, use &man.bsdlabel.8;: &prompt.root; /sbin/bsdlabel -B -w /dev/fd0 fd1440 The floppy is now ready to be high-level formatted with a file system. The floppy's file system can be either UFS or FAT, where FAT is generally a better choice for floppies. To format the floppy with FAT, issue: &prompt.root; /sbin/newfs_msdos /dev/fd0 The disk is now ready for use. To use the floppy, mount it with &man.mount.msdosfs.8;. One can also install and use emulators/mtools from the Ports Collection. Backup Basics Implementing a backup plan is essential in order to have the ability to recover from disk failure, accidental file deletion, random file corruption, or complete machine destruction, including destruction of on-site backups. The backup type and schedule will vary, depending upon the importance of the data, the granularity needed for file restores, and the amount of acceptable downtime. Some possible backup techniques include: Archives of the whole system, backed up onto permanent, off-site media. This provides protection against all of the problems listed above, but is slow and inconvenient to restore from, especially for non-privileged users. File system snapshots, which are useful for restoring deleted files or previous versions of files. Copies of whole file systems or disks which are synchronized with another system on the network using a scheduled net/rsync. Hardware or software RAID, which minimizes or avoids downtime when a disk fails. Typically, a mix of backup techniques is used. For example, one could create a schedule to automate a weekly, full system backup that is stored off-site and to supplement this backup with hourly ZFS snapshots. In addition, one could make a manual backup of individual directories or files before making file edits or deletions. This section describes some of the utilities which can be used to create and manage backups on a &os; system. File System Backups backup software dump / restore dump restore The traditional &unix; programs for backing up a file system are &man.dump.8;, which creates the backup, and &man.restore.8;, which restores the backup. These utilities work at the disk block level, below the abstractions of the files, links, and directories that are created by file systems. Unlike other backup software, dump backs up an entire file system and is unable to backup only part of a file system or a directory tree that spans multiple file systems. Instead of writing files and directories, dump writes the raw data blocks that comprise files and directories. If dump is used on the root directory, it will not back up /home, /usr or many other directories since these are typically mount points for other file systems or symbolic links into those file systems. When used to restore data, restore stores temporary files in /tmp/ by default. When using a recovery disk with a small /tmp, set TMPDIR to a directory with more free space in order for the restore to succeed. When using dump, be aware that some quirks remain from its early days in Version 6 of AT&T &unix;,circa 1975. The default parameters assume a backup to a 9-track tape, rather than to another type of media or to the high-density tapes available today. These defaults must be overridden on the command line. .rhosts It is possible to backup a file system across the network to a another system or to a tape drive attached to another computer. While the &man.rdump.8; and &man.rrestore.8; utilities can be used for this purpose, they are not considered to be secure. Instead, one can use dump and restore in a more secure fashion over an SSH connection. This example creates a full, compressed backup of /usr and sends the backup file to the specified host over a SSH connection. Using <command>dump</command> over <application>ssh</application> &prompt.root; /sbin/dump -0uan -f - /usr | gzip -2 | ssh -c blowfish \ targetuser@targetmachine.example.com dd of=/mybigfiles/dump-usr-l0.gz This example sets RSH in order to write the backup to a tape drive on a remote system over a SSH connection: Using <command>dump</command> over <application>ssh</application> with <envar>RSH</envar> Set &prompt.root; env RSH=/usr/bin/ssh /sbin/dump -0uan -f targetuser@targetmachine.example.com:/dev/sa0 /usr Directory Backups backup software tar Several built-in utilities are available for backing up and restoring specified files and directories as needed. A good choice for making a backup of all of the files in a directory is &man.tar.1;. This utility dates back to Version 6 of AT&T &unix; and by default assumes a recursive backup to a local tape device. Switches can be used to instead specify the name of a backup file. tar This example creates a compressed backup of the current directory and saves it to /tmp/mybackup.tgz. When creating a backup file, make sure that the backup is not saved to the same directory that is being backed up. Backing Up the Current Directory with <command>tar</command> &prompt.root; tar czvf /tmp/mybackup.tgz . To restore the entire backup, cd into the directory to restore into and specify the name of the backup. Note that this will overwrite any newer versions of files in the restore directory. When in doubt, restore to a temporary directory or specify the name of the file within the backup to restore. Restoring Up the Current Directory with <command>tar</command> &prompt.root; tar xzvf /tmp/mybackup.tgz There are dozens of available switches which are described in &man.tar.1;. This utility also supports the use of exclude patterns to specify which files should not be included when backing up the specified directory or restoring files from a backup. backup software cpio To create a backup using a specified list of files and directories, &man.cpio.1; is a good choice. Unlike tar, cpio does not know how to walk the directory tree and it must be provided the list of files to backup. For example, a list of files can be created using ls or find. This example creates a recursive listing of the current directory which is then piped to cpio in order to create an output backup file named /tmp/mybackup.cpio. Using <command>ls</command> and <command>cpio</command> to Make a Recursive Backup of the Current Directory &prompt.root; ls -R | cpio -ovF /tmp/mybackup.cpio backup software pax pax POSIX IEEE A backup utility which tries to bridge the features provided by tar and cpio is &man.pax.1;. Over the years, the various versions of tar and cpio became slightly incompatible. &posix; created pax which attempts to read and write many of the various cpio and tar formats, plus new formats of its own. The pax equivalent to the previous examples would be: Backing Up the Current Directory with <command>pax</command> &prompt.root; pax -wf /tmp/mybackup.pax . Using Data Tapes for Backups tape media While tape technology has continued to evolve, modern backup systems tend to combine off-site backups with local removable media. &os; supports any tape drive that uses SCSI, such as LTO or DAT. There is limited support for SATA and USB tape drives. For SCSI tape devices, &os; uses the &man.sa.4; driver and the /dev/sa0, /dev/nsa0, and /dev/esa0 devices. The physical device name is /dev/sa0. When /dev/nsa0 is used, the backup application will not rewind the tape after writing a file, which allows writing more than one file to a tape. Using /dev/esa0 ejects the tape after the device is closed. In &os;, mt is used to control operations of the tape drive, such as seeking through files on a tape or writing tape control marks to the tape. For example, the first three files on a tape can be preserved by skipping past them before writing a new file: &prompt.root; mt -f /dev/nsa0 fsf 3 This utility supports many operations. Refer to &man.mt.1; for details. To write a single file to tape using tar, specify the name of the tape device and the file to backup: &prompt.root; tar cvf /dev/sa0 file To recover files from a tar archive on tape into the current directory: &prompt.root; tar xvf /dev/sa0 To backup a UFS file system, use dump. This examples backs up /usr without rewinding the tape when finished: &prompt.root; dump -0aL -b64 -f /dev/nsa0 /usr To interactively restore files from a dump file on tape into the current directory: &prompt.root; restore -i -f /dev/nsa0 Third-Party Backup Utilities backup software The &os; Ports Collection provides many third-party utilities which can be used to schedule the creation of backups, simplify tape backup, and make backups easier and more convenient. Many of these applications are client/server based and can be used to automate the backups of a single system or all of the computers in a network. Popular utilities include Amanda, Bacula, rsync, and duplicity. Emergency Recovery In addition to regular backups, it is recommended to perform the following steps as part of an emergency preparedness plan. bsdlabel Create a print copy of the output of the following commands: gpart show more /etc/fstab dmesg livefs CD Store this printout and a copy of the installation media in a secure location. Should an emergency restore be needed, boot into the installation media and select Live CD to access a rescue shell. This rescue mode can be used to view the current state of the system, and if needed, to reformat disks and restore data from backups. The installation media for &os;/&arch.i386; &rel2.current;-RELEASE does not include a rescue shell. For this version, instead download and burn a Livefs CD image from ftp://ftp.FreeBSD.org/pub/FreeBSD/releases/&arch.i386;/ISO-IMAGES/&rel2.current;/&os;-&rel2.current;-RELEASE-&arch.i386;-livefs.iso. Next, test the rescue shell and the backups. Make notes of the procedure. Store these notes with the media, the printouts, and the backups. These notes may prevent the inadvertent destruction of the backups while under the stress of performing an emergency recovery. For an added measure of security, store the latest backup at a remote location which is physically separated from the computers and disk drives by a significant distance. Memory Disks Marc Fonvieille Reorganized and enhanced by In addition to physical disks, &os; also supports the creation and use of memory disks. One possible use for a memory disk is to access the contents of an ISO file system without the overhead of first burning it to a CD or DVD, then mounting the CD/DVD media. In &os;, the &man.md.4; driver is used to provide support for memory disks. The GENERIC kernel includes this driver. When using a custom kernel configuration file, ensure it includes this line: device md Attaching and Detaching Existing Images disks memory To mount an existing file system image, use mdconfig to specify the name of the ISO file and a free unit number. Then, refer to that unit number to mount it on an existing mount point. Once mounted, the files in the ISO will appear in the mount point. This example attaches diskimage.iso to the memory device /dev/md0 then mounts that memory device on /mnt: &prompt.root; mdconfig -f diskimage.iso -u 0 &prompt.root; mount /dev/md0 /mnt If a unit number is not specified with , mdconfig will automatically allocate an unused memory device and output the name of the allocated unit, such as md4. Refer to &man.mdconfig.8; for more details about this command and its options. disks detaching a memory disk When a memory disk is no longer in use, its resources should be released back to the system. First, unmount the file system, then use mdconfig to detach the disk from the system and release its resources. To continue this example: &prompt.root; umount /mnt &prompt.root; mdconfig -d -u 0 To determine if any memory disks are still attached to the system, type mdconfig -l. Creating a File- or Memory-Backed Memory Disk disks memory file system &os; also supports memory disks where the storage to use is allocated from either a hard disk or an area of memory. The first method is commonly referred to as a file-backed file system and the second method as a memory-backed file system. Both types can be created using mdconfig. To create a new memory-backed file system, specify a type of swap and the size of the memory disk to create. Then, format the memory disk with a file system and mount as usual. This example creates a 5M memory disk on unit 1. That memory disk is then formatted with the UFS file system before it is mounted: &prompt.root; mdconfig -a -t swap -s 5m -u 1 &prompt.root; newfs -U md1 /dev/md1: 5.0MB (10240 sectors) block size 16384, fragment size 2048 using 4 cylinder groups of 1.27MB, 81 blks, 192 inodes. with soft updates super-block backups (for fsck -b #) at: 160, 2752, 5344, 7936 &prompt.root; mount /dev/md1 /mnt &prompt.root; df /mnt Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/md1 4718 4 4338 0% /mnt To create a new file-backed memory disk, first allocate an area of disk to use. This example creates an empty 5K file named newimage: &prompt.root; dd if=/dev/zero of=newimage bs=1k count=5k 5120+0 records in 5120+0 records out Next, attach that file to a memory disk, label the memory disk and format it with the UFS file system, mount the memory disk, and verify the size of the file-backed disk: &prompt.root; mdconfig -f newimage -u 0 &prompt.root; bsdlabel -w md0 auto &prompt.root; newfs md0a /dev/md0a: 5.0MB (10224 sectors) block size 16384, fragment size 2048 using 4 cylinder groups of 1.25MB, 80 blks, 192 inodes. super-block backups (for fsck -b #) at: 160, 2720, 5280, 7840 &prompt.root; mount /dev/md0a /mnt &prompt.root; df /mnt Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/md0a 4710 4 4330 0% /mnt It takes several commands to create a file- or memory-backed file system using mdconfig. &os; also comes with mdmfs which automatically configures a memory disk, formats it with the UFS file system, and mounts it. For example, after creating newimage with dd, this one command is equivalent to running the bsdlabel, newfs, and mount commands shown above: &prompt.root; mdmfs -F newimage -s 5m md0 /mnt To instead create a new memory-based memory disk with mdmfs, use this one command: &prompt.root; mdmfs -s 5m md1 /mnt If the unit number is not specified, mdmfs will automatically select an unused memory device. For more details about mdmfs, refer to &man.mdmfs.8;. File System Snapshots Tom Rhodes Contributed by file systems snapshots &os; offers a feature in conjunction with Soft Updates: file system snapshots. UFS snapshots allow a user to create images of specified file systems, and treat them as a file. Snapshot files must be created in the file system that the action is performed on, and a user may create no more than 20 snapshots per file system. Active snapshots are recorded in the superblock so they are persistent across unmount and remount operations along with system reboots. When a snapshot is no longer required, it can be removed using &man.rm.1;. While snapshots may be removed in any order, all the used space may not be acquired because another snapshot will possibly claim some of the released blocks. The un-alterable file flag is set by &man.mksnap.ffs.8; after initial creation of a snapshot file. &man.unlink.1; makes an exception for snapshot files since it allows them to be removed. Snapshots are created using &man.mount.8;. To place a snapshot of /var in the file /var/snapshot/snap, use the following command: &prompt.root; mount -u -o snapshot /var/snapshot/snap /var Alternatively, use &man.mksnap.ffs.8; to create the snapshot: &prompt.root; mksnap_ffs /var /var/snapshot/snap One can find snapshot files on a file system, such as /var, using &man.find.1;: &prompt.root; find /var -flags snapshot Once a snapshot has been created, it has several uses: Some administrators will use a snapshot file for backup purposes, because the snapshot can be transferred to CDs or tape. The file system integrity checker, &man.fsck.8;, may be run on the snapshot. Assuming that the file system was clean when it was mounted, this should always provide a clean and unchanging result. Running &man.dump.8; on the snapshot will produce a dump file that is consistent with the file system and the timestamp of the snapshot. &man.dump.8; can also take a snapshot, create a dump image, and then remove the snapshot in one command by using . The snapshot can be mounted as a frozen image of the file system. To &man.mount.8; the snapshot /var/snapshot/snap run: &prompt.root; mdconfig -a -t vnode -o readonly -f /var/snapshot/snap -u 4 &prompt.root; mount -r /dev/md4 /mnt The frozen /var is now available through /mnt. Everything will initially be in the same state it was during the snapshot creation time. The only exception is that any earlier snapshots will appear as zero length files. To unmount the snapshot, use: &prompt.root; umount /mnt &prompt.root; mdconfig -d -u 4 For more information about and file system snapshots, including technical papers, visit Marshall Kirk McKusick's website at http://www.mckusick.com/. Disk Quotas accounting disk space disk quotas Disk quotas can be used to limit the amount of disk space or the number of files a user or members of a group may allocate on a per-file system basis. This prevents one user or group of users from consuming all of the available disk space. This section describes how to configure disk quotas for the UFS file system. To configure quotas on the ZFS file system, refer to Enabling Disk Quotas To determine if the &os; kernel provides support for disk quotas: &prompt.user; sysctl kern.features.ufs_quota kern.features.ufs_quota: 1 In this example, the 1 indicates quota support. If the value is instead 0, add the following line to a custom kernel configuration file and rebuild the kernel using the instructions in : options QUOTA Next, enable disk quotas in /etc/rc.conf: quota_enable="YES" disk quotas checking Normally on bootup, the quota integrity of each file system is checked by &man.quotacheck.8;. This program insures that the data in the quota database properly reflects the data on the file system. This is a time consuming process that will significantly affect the time the system takes to boot. To skip this step, add this variable to /etc/rc.conf: check_quotas="NO" Finally, edit /etc/fstab to enable disk quotas on a per-file system basis. To enable per-user quotas on a file system, add to the options field in the /etc/fstab entry for the file system to enable quotas on. For example: /dev/da1s2g /home ufs rw,userquota 1 2 To enable group quotas, use instead. To enable both user and group quotas, separate the options with a comma: /dev/da1s2g /home ufs rw,userquota,groupquota 1 2 By default, quota files are stored in the root directory of the file system as quota.user and quota.group. Refer to &man.fstab.5; for more information. Specifying an alternate location for the quota files is not recommended. Once the configuration is complete, reboot the system and /etc/rc will automatically run the appropriate commands to create the initial quota files for all of the quotas enabled in /etc/fstab. In the normal course of operations, there should be no need to manually run &man.quotacheck.8;, &man.quotaon.8;, or &man.quotaoff.8;. However, one should read these manual pages to be familiar with their operation. Setting Quota Limits disk quotas limits To verify that quotas are enabled, run: &prompt.root; quota -v There should be a one line summary of disk usage and current quota limits for each file system that quotas are enabled on. The system is now ready to be assigned quota limits with edquota. Several options are available to enforce limits on the amount of disk space a user or group may allocate, and how many files they may create. Allocations can be limited based on disk space (block quotas), number of files (inode quotas), or a combination of both. Each limit is further broken down into two categories: hard and soft limits. hard limit A hard limit may not be exceeded. Once a user reaches a hard limit, no further allocations can be made on that file system by that user. For example, if the user has a hard limit of 500 kbytes on a file system and is currently using 490 kbytes, the user can only allocate an additional 10 kbytes. Attempting to allocate an additional 11 kbytes will fail. soft limit Soft limits can be exceeded for a limited amount of time, known as the grace period, which is one week by default. If a user stays over their limit longer than the grace period, the soft limit turns into a hard limit and no further allocations are allowed. When the user drops back below the soft limit, the grace period is reset. In the following example, the quota for the test account is being edited. When edquota is invoked, the editor specified by EDITOR is opened in order to edit the quota limits. The default editor is set to vi. &prompt.root; edquota -u test Quotas for user test: /usr: kbytes in use: 65, limits (soft = 50, hard = 75) inodes in use: 7, limits (soft = 50, hard = 60) /usr/var: kbytes in use: 0, limits (soft = 50, hard = 75) inodes in use: 0, limits (soft = 50, hard = 60) There are normally two lines for each file system that has quotas enabled. One line represents the block limits and the other represents the inode limits. Change the value to modify the quota limit. For example, to raise the block limit on /usr to a soft limit of 500 and a hard limit of 600, change the values in that line as follows: /usr: kbytes in use: 65, limits (soft = 500, hard = 600) The new quota limits take effect upon exiting the editor. Sometimes it is desirable to set quota limits on a range of users. This can be done by first assigning the desired quota limit to a user. Then, use to duplicate that quota to a specified range of user IDs (UIDs). The following command will duplicate those quota limits for UIDs 10,000 through 19,999: &prompt.root; edquota -p test 10000-19999 For more information, refer to &man.edquota.8;. Checking Quota Limits and Disk Usage disk quotas checking To check individual user or group quotas and disk usage, use &man.quota.1;. A user may only examine their own quota and the quota of a group they are a member of. Only the superuser may view all user and group quotas. To get a summary of all quotas and disk usage for file systems with quotas enabled, use &man.repquota.8;. Normally, file systems that the user is not using any disk space on will not show in the output of quota, even if the user has a quota limit assigned for that file system. Use to display those file systems. The following is sample output from quota -v for a user that has quota limits on two file systems. Disk quotas for user test (uid 1002): Filesystem usage quota limit grace files quota limit grace /usr 65* 50 75 5days 7 50 60 /usr/var 0 50 75 0 50 60 grace period In this example, the user is currently 15 kbytes over the soft limit of 50 kbytes on /usr and has 5 days of grace period left. The asterisk * indicates that the user is currently over the quota limit. Quotas over NFS NFS Quotas are enforced by the quota subsystem on the NFS server. The &man.rpc.rquotad.8; daemon makes quota information available to quota on NFS clients, allowing users on those machines to see their quota statistics. On the NFS server, enable rpc.rquotad by removing the # from this line in /etc/inetd.conf: rquotad/1 dgram rpc/udp wait root /usr/libexec/rpc.rquotad rpc.rquotad Then, restart inetd: &prompt.root; service inetd restart Encrypting Disk Partitions Lucky Green Contributed by
shamrock@cypherpunks.to
disks encrypting &os; offers excellent online protections against unauthorized data access. File permissions and Mandatory Access Control (MAC) help prevent unauthorized users from accessing data while the operating system is active and the computer is powered up. However, the permissions enforced by the operating system are irrelevant if an attacker has physical access to a computer and can move the computer's hard drive to another system to copy and analyze the data. Regardless of how an attacker may have come into possession of a hard drive or powered-down computer, the GEOM-based cryptographic subsystems built into &os; are able to protect the data on the computer's file systems against even highly-motivated attackers with significant resources. Unlike encryption methods that encrypt individual files, the built-in gbde and geli utilities can be used to transparently encrypt entire file systems. No cleartext ever touches the hard drive's platter. This chapter demonstrates how to create an encrypted file system on &os;. It first demonstrates the process using gbde and then demonstrates the same example using geli. Disk Encryption with <application>gbde</application> The objective of the &man.gbde.4; facility is to provide a formidable challenge for an attacker to gain access to the contents of a cold storage device. However, if the computer is compromised while up and running and the storage device is actively attached, or the attacker has access to a valid passphrase, it offers no protection to the contents of the storage device. Thus, it is important to provide physical security while the system is running and to protect the passphrase used by the encryption mechanism. This facility provides several barriers to protect the data stored in each disk sector. It encrypts the contents of a disk sector using 128-bit AES in CBC mode. Each sector on the disk is encrypted with a different AES key. For more information on the cryptographic design, including how the sector keys are derived from the user-supplied passphrase, refer to &man.gbde.4;. &os; provides a kernel module for gbde which can be loaded with this command: &prompt.root; kldload geom_bde If using a custom kernel configuration file, ensure it contains this line: options GEOM_BDE The following example demonstrates adding a new hard drive to a system that will hold a single encrypted partition that will be mounted as /private. Encrypting a Partition with <application>gbde</application> Add the New Hard Drive Install the new drive to the system as explained in . For the purposes of this example, a new hard drive partition has been added as /dev/ad4s1c and /dev/ad0s1* represents the existing standard &os; partitions. &prompt.root; ls /dev/ad* /dev/ad0 /dev/ad0s1b /dev/ad0s1e /dev/ad4s1 /dev/ad0s1 /dev/ad0s1c /dev/ad0s1f /dev/ad4s1c /dev/ad0s1a /dev/ad0s1d /dev/ad4 Create a Directory to Hold <command>gbde</command> Lock Files &prompt.root; mkdir /etc/gbde The gbde lock file contains information that gbde requires to access encrypted partitions. Without access to the lock file, gbde will not be able to decrypt the data contained in the encrypted partition without significant manual intervention which is not supported by the software. Each encrypted partition uses a separate lock file. Initialize the <command>gbde</command> Partition A gbde partition must be initialized before it can be used. This initialization needs to be performed only once. This command will open the default editor, in order to set various configuration options in a template. For use with the UFS file system, set the sector_size to 2048: - &prompt.root; gbde init /dev/ad4s1c -i -L /etc/gbde/ad4s1c.lock -# $FreeBSD: src/sbin/gbde/template.txt,v 1.1.36.1 2009/08/03 08:13:06 kensmith Exp $ + &prompt.root; gbde init /dev/ad4s1c -i -L /etc/gbde/ad4s1c.lock +# $FreeBSD: src/sbin/gbde/template.txt,v 1.1.36.1 2009/08/03 08:13:06 kensmith Exp $ # # Sector size is the smallest unit of data which can be read or written. # Making it too small decreases performance and decreases available space. # Making it too large may prevent filesystems from working. 512 is the # minimum and always safe. For UFS, use the fragment size # sector_size = 2048 [...] Once the edit is saved, the user will be asked twice to type the passphrase used to secure the data. The passphrase must be the same both times. The ability of gbde to protect data depends entirely on the quality of the passphrase. For tips on how to select a secure passphrase that is easy to remember, see http://world.std.com/~reinhold/diceware.htm. This initialization creates a lock file for the gbde partition. In this example, it is stored as /etc/gbde/ad4s1c.lock. Lock files must end in .lock in order to be correctly detected by the /etc/rc.d/gbde start up script. Lock files must be backed up together with the contents of any encrypted partitions. Without the lock file, the legitimate owner will be unable to access the data on the encrypted partition. Attach the Encrypted Partition to the Kernel &prompt.root; gbde attach /dev/ad4s1c -l /etc/gbde/ad4s1c.lock This command will prompt to input the passphrase that was selected during the initialization of the encrypted partition. The new encrypted device will appear in /dev as /dev/device_name.bde: &prompt.root; ls /dev/ad* /dev/ad0 /dev/ad0s1b /dev/ad0s1e /dev/ad4s1 /dev/ad0s1 /dev/ad0s1c /dev/ad0s1f /dev/ad4s1c /dev/ad0s1a /dev/ad0s1d /dev/ad4 /dev/ad4s1c.bde Create a File System on the Encrypted Device Once the encrypted device has been attached to the kernel, a file system can be created on the device. This example creates a UFS file system with soft updates enabled. Be sure to specify the partition which has a *.bde extension: &prompt.root; newfs -U /dev/ad4s1c.bde Mount the Encrypted Partition Create a mount point and mount the encrypted file system: &prompt.root; mkdir /private &prompt.root; mount /dev/ad4s1c.bde /private Verify That the Encrypted File System is Available The encrypted file system should now be visible and available for use: &prompt.user; df -H Filesystem Size Used Avail Capacity Mounted on /dev/ad0s1a 1037M 72M 883M 8% / /devfs 1.0K 1.0K 0B 100% /dev /dev/ad0s1f 8.1G 55K 7.5G 0% /home /dev/ad0s1e 1037M 1.1M 953M 0% /tmp /dev/ad0s1d 6.1G 1.9G 3.7G 35% /usr /dev/ad4s1c.bde 150G 4.1K 138G 0% /private After each boot, any encrypted file systems must be manually re-attached to the kernel, checked for errors, and mounted, before the file systems can be used. To configure these steps, add the following lines to /etc/rc.conf: gbde_autoattach_all="YES" gbde_devices="ad4s1c" gbde_lockdir="/etc/gbde" This requires that the passphrase be entered at the console at boot time. After typing the correct passphrase, the encrypted partition will be mounted automatically. Additional gbde boot options are available and listed in &man.rc.conf.5;. sysinstall is incompatible with gbde-encrypted devices. All *.bde devices must be detached from the kernel before starting sysinstall or it will crash during its initial probing for devices. To detach the encrypted device used in the example, use the following command: &prompt.root; gbde detach /dev/ad4s1c Disk Encryption with <command>geli</command> Daniel Gerzo Contributed by An alternative cryptographic GEOM class is available using geli. This control utility adds some features and uses a different scheme for doing cryptographic work. It provides the following features: Utilizes the &man.crypto.9; framework and automatically uses cryptographic hardware when it is available. Supports multiple cryptographic algorithms such as AES, Blowfish, and 3DES. Allows the root partition to be encrypted. The passphrase used to access the encrypted root partition will be requested during system boot. Allows the use of two independent keys. It is fast as it performs simple sector-to-sector encryption. Allows backup and restore of master keys. If a user destroys their keys, it is still possible to get access to the data by restoring keys from the backup. Allows a disk to attach with a random, one-time key which is useful for swap partitions and temporary file systems. More features and usage examples can be found in &man.geli.8;. The following example describes how to generate a key file which will be used as part of the master key for the encrypted provider mounted under /private. The key file will provide some random data used to encrypt the master key. The master key will also be protected by a passphrase. The provider's sector size will be 4kB. The example describes how to attach to the geli provider, create a file system on it, mount it, work with it, and finally, how to detach it. Encrypting a Partition with <command>geli</command> Load <command>geli</command> Support Support for geli is available as a loadable kernel module. To configure the system to automatically load the module at boot time, add the following line to /boot/loader.conf: geom_eli_load="YES" To load the kernel module now: &prompt.root; kldload geom_eli For a custom kernel, ensure the kernel configuration file contains these lines: options GEOM_ELI device crypto Generate the Master Key The following commands generate a master key (/root/da2.key) that is protected with a passphrase. The data source for the key file is /dev/random and the sector size of the provider (/dev/da2.eli) is 4kB as a bigger sector size provides better performance: &prompt.root; dd if=/dev/random of=/root/da2.key bs=64 count=1 &prompt.root; geli init -s 4096 -K /root/da2.key /dev/da2 Enter new passphrase: Reenter new passphrase: It is not mandatory to use both a passphrase and a key file as either method of securing the master key can be used in isolation. If the key file is given as -, standard input will be used. For example, this command generates three key files: &prompt.root; cat keyfile1 keyfile2 keyfile3 | geli init -K - /dev/da2 Attach the Provider with the Generated Key To attach the provider, specify the key file, the name of the disk, and the passphrase: &prompt.root; geli attach -k /root/da2.key /dev/da2 Enter passphrase: This creates a new device with an .eli extension: &prompt.root; ls /dev/da2* /dev/da2 /dev/da2.eli Create the New File System Next, format the device with the UFS file system and mount it on an existing mount point: &prompt.root; dd if=/dev/random of=/dev/da2.eli bs=1m &prompt.root; newfs /dev/da2.eli &prompt.root; mount /dev/da2.eli /private The encrypted file system should now be available for use: &prompt.root; df -H Filesystem Size Used Avail Capacity Mounted on /dev/ad0s1a 248M 89M 139M 38% / /devfs 1.0K 1.0K 0B 100% /dev /dev/ad0s1f 7.7G 2.3G 4.9G 32% /usr /dev/ad0s1d 989M 1.5M 909M 0% /tmp /dev/ad0s1e 3.9G 1.3G 2.3G 35% /var /dev/da2.eli 150G 4.1K 138G 0% /private Once the work on the encrypted partition is done, and the /private partition is no longer needed, it is prudent to put the device into cold storage by unmounting and detaching the geli encrypted partition from the kernel: &prompt.root; umount /private &prompt.root; geli detach da2.eli A rc.d script is provided to simplify the mounting of geli-encrypted devices at boot time. For this example, add these lines to /etc/rc.conf: geli_devices="da2" geli_da2_flags="-k /root/da2.key" This configures /dev/da2 as a geli provider with a master key of /root/da2.key. The system will automatically detach the provider from the kernel before the system shuts down. During the startup process, the script will prompt for the passphrase before attaching the provider. Other kernel messages might be shown before and after the password prompt. If the boot process seems to stall, look carefully for the password prompt among the other messages. Once the correct passphrase is entered, the provider is attached. The file system is then mounted, typically by an entry in /etc/fstab. Refer to for instructions on how to configure a file system to mount at boot time. Encrypting Swap Christian Brueffer Written by swap encrypting Like the encryption of disk partitions, encryption of swap space is used to protect sensitive information. Consider an application that deals with passwords. As long as these passwords stay in physical memory, they are not written to disk and will be cleared after a reboot. However, if &os; starts swapping out memory pages to free space, the passwords may be written to the disk unencrypted. Encrypting swap space can be a solution for this scenario. This section demonstrates how to configure an encrypted swap partition using &man.gbde.8; or &man.geli.8; encryption. It assumes that /dev/ada0s1b is the swap partition. Configuring Encrypted Swap Swap partitions are not encrypted by default and should be cleared of any sensitive data before continuing. To overwrite the current swap partition with random garbage, execute the following command: &prompt.root; dd if=/dev/random of=/dev/ada0s1b bs=1m To encrypt the swap partition using &man.gbde.8;, add the .bde suffix to the swap line in /etc/fstab: # Device Mountpoint FStype Options Dump Pass# /dev/ada0s1b.bde none swap sw 0 0 To instead encrypt the swap partition using &man.geli.8;, use the .eli suffix: # Device Mountpoint FStype Options Dump Pass# /dev/ada0s1b.eli none swap sw 0 0 By default, &man.geli.8; uses the AES algorithm with a key length of 128 bits. Normally the default settings will suffice. If desired, these defaults can be altered in the options field in /etc/fstab. The possible flags are: aalgo Data integrity verification algorithm used to ensure that the encrypted data has not been tampered with. See &man.geli.8; for a list of supported algorithms. ealgo Encryption algorithm used to protect the data. See &man.geli.8; for a list of supported algorithms. keylen The length of the key used for the encryption algorithm. See &man.geli.8; for the key lengths that are supported by each encryption algorithm. sectorsize The size of the blocks data is broken into before it is encrypted. Larger sector sizes increase performance at the cost of higher storage overhead. The recommended size is 4096 bytes. This example configures an encryped swap partition using the Blowfish algorithm with a key length of 128 bits and a sectorsize of 4 kilobytes: # Device Mountpoint FStype Options Dump Pass# /dev/ada0s1b.eli none swap sw,ealgo=blowfish,keylen=128,sectorsize=4096 0 0 Encrypted Swap Verification Once the system has rebooted, proper operation of the encrypted swap can be verified using swapinfo. If &man.gbde.8; is being used: &prompt.user; swapinfo Device 1K-blocks Used Avail Capacity /dev/ada0s1b.bde 542720 0 542720 0% If &man.geli.8; is being used: &prompt.user; swapinfo Device 1K-blocks Used Avail Capacity /dev/ada0s1b.eli 542720 0 542720 0% Highly Available Storage (<acronym>HAST</acronym>) Daniel Gerzo Contributed by Freddie Cash With inputs from Pawel Jakub Dawidek Michael W. Lucas Viktor Petersson HAST high availability High availability is one of the main requirements in serious business applications and highly-available storage is a key component in such environments. In &os;, the Highly Available STorage (HAST) framework allows transparent storage of the same data across several physically separated machines connected by a TCP/IP network. HAST can be understood as a network-based RAID1 (mirror), and is similar to the DRBD® storage system used in the GNU/&linux; platform. In combination with other high-availability features of &os; like CARP, HAST makes it possible to build a highly-available storage cluster that is resistant to hardware failures. The following are the main features of HAST: Can be used to mask I/O errors on local hard drives. File system agnostic as it works with any file system supported by &os;. Efficient and quick resynchronization as only the blocks that were modified during the downtime of a node are synchronized. Can be used in an already deployed environment to add additional redundancy. Together with CARP, Heartbeat, or other tools, it can be used to build a robust and durable storage system. After reading this section, you will know: What HAST is, how it works, and which features it provides. How to set up and use HAST on &os;. How to integrate CARP and &man.devd.8; to build a robust storage system. Before reading this section, you should: Understand &unix; and &os; basics (). Know how to configure network interfaces and other core &os; subsystems (). Have a good understanding of &os; networking (). The HAST project was sponsored by The &os; Foundation with support from http://www.omc.net/ and http://www.transip.nl/. HAST Operation HAST provides synchronous block-level replication between two physical machines: the primary, also known as the master node, and the secondary, or slave node. These two machines together are referred to as a cluster. Since HAST works in a primary-secondary configuration, it allows only one of the cluster nodes to be active at any given time. The primary node, also called active, is the one which will handle all the I/O requests to HAST-managed devices. The secondary node is automatically synchronized from the primary node. The physical components of the HAST system are the local disk on primary node, and the disk on the remote, secondary node. HAST operates synchronously on a block level, making it transparent to file systems and applications. HAST provides regular GEOM providers in /dev/hast/ for use by other tools or applications. There is no difference between using HAST-provided devices and raw disks or partitions. Each write, delete, or flush operation is sent to both the local disk and to the remote disk over TCP/IP. Each read operation is served from the local disk, unless the local disk is not up-to-date or an I/O error occurs. In such cases, the read operation is sent to the secondary node. HAST tries to provide fast failure recovery. For this reason, it is important to reduce synchronization time after a node's outage. To provide fast synchronization, HAST manages an on-disk bitmap of dirty extents and only synchronizes those during a regular synchronization, with an exception of the initial sync. There are many ways to handle synchronization. HAST implements several replication modes to handle different synchronization methods: memsync: This mode reports a write operation as completed when the local write operation is finished and when the remote node acknowledges data arrival, but before actually storing the data. The data on the remote node will be stored directly after sending the acknowledgement. This mode is intended to reduce latency, but still provides good reliability. This mode is the default. fullsync: This mode reports a write operation as completed when both the local write and the remote write complete. This is the safest and the slowest replication mode. async: This mode reports a write operation as completed when the local write completes. This is the fastest and the most dangerous replication mode. It should only be used when replicating to a distant node where latency is too high for other modes. HAST Configuration The HAST framework consists of several components: The &man.hastd.8; daemon which provides data synchronization. When this daemon is started, it will automatically load geom_gate.ko. The userland management utility, &man.hastctl.8;. The &man.hast.conf.5; configuration file. This file must exist before starting hastd. Users who prefer to statically build GEOM_GATE support into the kernel should add this line to the custom kernel configuration file, then rebuild the kernel using the instructions in : options GEOM_GATE The following example describes how to configure two nodes in master-slave/primary-secondary operation using HAST to replicate the data between the two. The nodes will be called hasta, with an IP address of 172.16.0.1, and hastb, with an IP address of 172.16.0.2. Both nodes will have a dedicated hard drive /dev/ad6 of the same size for HAST operation. The HAST pool, sometimes referred to as a resource or the GEOM provider in /dev/hast/, will be called test. Configuration of HAST is done using /etc/hast.conf. This file should be identical on both nodes. The simplest configuration is: resource test { on hasta { local /dev/ad6 remote 172.16.0.2 } on hastb { local /dev/ad6 remote 172.16.0.1 } } For more advanced configuration, refer to &man.hast.conf.5;. It is also possible to use host names in the remote statements if the hosts are resolvable and defined either in /etc/hosts or in the local DNS. Once the configuration exists on both nodes, the HAST pool can be created. Run these commands on both nodes to place the initial metadata onto the local disk and to start &man.hastd.8;: &prompt.root; hastctl create test &prompt.root; service hastd onestart It is not possible to use GEOM providers with an existing file system or to convert an existing storage to a HAST-managed pool. This procedure needs to store some metadata on the provider and there will not be enough required space available on an existing provider. A HAST node's primary or secondary role is selected by an administrator, or software like Heartbeat, using &man.hastctl.8;. On the primary node, hasta, issue this command: &prompt.root; hastctl role primary test Run this command on the secondary node, hastb: &prompt.root; hastctl role secondary test Verify the result by running hastctl on each node: &prompt.root; hastctl status test Check the status line in the output. If it says degraded, something is wrong with the configuration file. It should say complete on each node, meaning that the synchronization between the nodes has started. The synchronization completes when hastctl status reports 0 bytes of dirty extents. The next step is to create a file system on the GEOM provider and mount it. This must be done on the primary node. Creating the file system can take a few minutes, depending on the size of the hard drive. This example creates a UFS file system on /dev/hast/test: &prompt.root; newfs -U /dev/hast/test &prompt.root; mkdir /hast/test &prompt.root; mount /dev/hast/test /hast/test Once the HAST framework is configured properly, the final step is to make sure that HAST is started automatically during system boot. Add this line to /etc/rc.conf: hastd_enable="YES" Failover Configuration The goal of this example is to build a robust storage system which is resistant to the failure of any given node. If the primary node fails, the secondary node is there to take over seamlessly, check and mount the file system, and continue to work without missing a single bit of data. To accomplish this task, the Common Address Redundancy Protocol (CARP) is used to provide for automatic failover at the IP layer. CARP allows multiple hosts on the same network segment to share an IP address. Set up CARP on both nodes of the cluster according to the documentation available in . In this example, each node will have its own management IP address and a shared IP address of 172.16.0.254. The primary HAST node of the cluster must be the master CARP node. The HAST pool created in the previous section is now ready to be exported to the other hosts on the network. This can be accomplished by exporting it through NFS or Samba, using the shared IP address 172.16.0.254. The only problem which remains unresolved is an automatic failover should the primary node fail. In the event of CARP interfaces going up or down, the &os; operating system generates a &man.devd.8; event, making it possible to watch for state changes on the CARP interfaces. A state change on the CARP interface is an indication that one of the nodes failed or came back online. These state change events make it possible to run a script which will automatically handle the HAST failover. To catch state changes on the CARP interfaces, add this configuration to /etc/devd.conf on each node: notify 30 { match "system" "IFNET"; match "subsystem" "carp0"; match "type" "LINK_UP"; action "/usr/local/sbin/carp-hast-switch master"; }; notify 30 { match "system" "IFNET"; match "subsystem" "carp0"; match "type" "LINK_DOWN"; action "/usr/local/sbin/carp-hast-switch slave"; }; If the systems are running &os; 10 or higher, replace carp0 with the name of the CARP-configured interface. Restart &man.devd.8; on both nodes to put the new configuration into effect: &prompt.root; service devd restart When the specified interface state changes by going up or down , the system generates a notification, allowing the &man.devd.8; subsystem to run the specified automatic failover script, /usr/local/sbin/carp-hast-switch. For further clarification about this configuration, refer to &man.devd.conf.5;. Here is an example of an automated failover script: #!/bin/sh # Original script by Freddie Cash <fjwcash@gmail.com> # Modified by Michael W. Lucas <mwlucas@BlackHelicopters.org> # and Viktor Petersson <vpetersson@wireload.net> # The names of the HAST resources, as listed in /etc/hast.conf resources="test" # delay in mounting HAST resource after becoming master # make your best guess delay=3 # logging log="local0.debug" name="carp-hast" # end of user configurable stuff case "$1" in master) logger -p $log -t $name "Switching to primary provider for ${resources}." sleep ${delay} # Wait for any "hastd secondary" processes to stop for disk in ${resources}; do while $( pgrep -lf "hastd: ${disk} \(secondary\)" > /dev/null 2>&1 ); do sleep 1 done # Switch role for each disk hastctl role primary ${disk} if [ $? -ne 0 ]; then logger -p $log -t $name "Unable to change role to primary for resource ${disk}." exit 1 fi done # Wait for the /dev/hast/* devices to appear for disk in ${resources}; do for I in $( jot 60 ); do [ -c "/dev/hast/${disk}" ] && break sleep 0.5 done if [ ! -c "/dev/hast/${disk}" ]; then logger -p $log -t $name "GEOM provider /dev/hast/${disk} did not appear." exit 1 fi done logger -p $log -t $name "Role for HAST resources ${resources} switched to primary." logger -p $log -t $name "Mounting disks." for disk in ${resources}; do mkdir -p /hast/${disk} fsck -p -y -t ufs /dev/hast/${disk} mount /dev/hast/${disk} /hast/${disk} done ;; slave) logger -p $log -t $name "Switching to secondary provider for ${resources}." # Switch roles for the HAST resources for disk in ${resources}; do if ! mount | grep -q "^/dev/hast/${disk} on " then else umount -f /hast/${disk} fi sleep $delay hastctl role secondary ${disk} 2>&1 if [ $? -ne 0 ]; then logger -p $log -t $name "Unable to switch role to secondary for resource ${disk}." exit 1 fi logger -p $log -t $name "Role switched to secondary for resource ${disk}." done ;; esac In a nutshell, the script takes these actions when a node becomes master: Promotes the HAST pool to primary on the other node. Checks the file system under the HAST pool. Mounts the pool. When a node becomes secondary: Unmounts the HAST pool. Degrades the HAST pool to secondary. This is just an example script which serves as a proof of concept. It does not handle all the possible scenarios and can be extended or altered in any way, for example, to start or stop required services. For this example, a standard UFS file system was used. To reduce the time needed for recovery, a journal-enabled UFS or ZFS file system can be used instead. More detailed information with additional examples can be found at http://wiki.FreeBSD.org/HAST. Troubleshooting HAST should generally work without issues. However, as with any other software product, there may be times when it does not work as supposed. The sources of the problems may be different, but the rule of thumb is to ensure that the time is synchronized between the nodes of the cluster. When troubleshooting HAST, the debugging level of &man.hastd.8; should be increased by starting hastd with -d. This argument may be specified multiple times to further increase the debugging level. Consider also using -F, which starts hastd in the foreground. Recovering from the Split-brain Condition Split-brain occurs when the nodes of the cluster are unable to communicate with each other, and both are configured as primary. This is a dangerous condition because it allows both nodes to make incompatible changes to the data. This problem must be corrected manually by the system administrator. The administrator must either decide which node has more important changes, or perform the merge manually. Then, let HAST perform full synchronization of the node which has the broken data. To do this, issue these commands on the node which needs to be resynchronized: &prompt.root; hastctl role init test &prompt.root; hastctl create test &prompt.root; hastctl role secondary test Index: head/en_US.ISO8859-1/books/handbook/mail/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/mail/chapter.xml (revision 50807) +++ head/en_US.ISO8859-1/books/handbook/mail/chapter.xml (revision 50808) @@ -1,1951 +1,1951 @@ Electronic Mail BillLloydOriginal work by JimMockRewritten by Synopsis email Electronic Mail, better known as email, is one of the most widely used forms of communication today. This chapter provides a basic introduction to running a mail server on &os;, as well as an introduction to sending and receiving email using &os;. For more complete coverage of this subject, refer to the books listed in . After reading this chapter, you will know: Which software components are involved in sending and receiving electronic mail. Where basic Sendmail configuration files are located in &os;. The difference between remote and local mailboxes. How to block spammers from illegally using a mail server as a relay. How to install and configure an alternate Mail Transfer Agent, replacing Sendmail. How to troubleshoot common mail server problems. How to set up the system to send mail only. How to use mail with a dialup connection. How to configure SMTP authentication for added security. How to install and use a Mail User Agent, such as mutt, to send and receive email. How to download mail from a remote POP or IMAP server. How to automatically apply filters and rules to incoming email. Before reading this chapter, you should: Properly set up a network connection (). Properly set up the DNS information for a mail host (). Know how to install additional third-party software (). Mail Components POP IMAP DNS mail server daemons Sendmail mail server daemons Postfix mail server daemons qmail mail server daemons Exim email receiving MX record mail host There are five major parts involved in an email exchange: the Mail User Agent (MUA), the Mail Transfer Agent (MTA), a mail host, a remote or local mailbox, and DNS. This section provides an overview of these components. Mail User Agent (MUA) The Mail User Agent (MUA) is an application which is used to compose, send, and receive emails. This application can be a command line program, such as the built-in mail utility or a third-party application from the Ports Collection, such as mutt, alpine, or elm. Dozens of graphical programs are also available in the Ports Collection, including Claws Mail, Evolution, and Thunderbird. Some organizations provide a web mail program which can be accessed through a web browser. More information about installing and using a MUA on &os; can be found in . Mail Transfer Agent (MTA) The Mail Transfer Agent (MTA) is responsible for receiving incoming mail and delivering outgoing mail. &os; ships with Sendmail as the default MTA, but it also supports numerous other mail server daemons, including Exim, Postfix, and qmail. Sendmail configuration is described in . If another MTA is installed using the Ports Collection, refer to its post-installation message for &os;-specific configuration details and the application's website for more general configuration instructions. Mail Host and Mailboxes The mail host is a server that is responsible for delivering and receiving mail for a host or a network. The mail host collects all mail sent to the domain and stores it either in the default mbox or the alternative Maildir format, depending on the configuration. Once mail has been stored, it may either be read locally using a MUA or remotely accessed and collected using protocols such as POP or IMAP. If mail is read locally, a POP or IMAP server does not need to be installed. To access mailboxes remotely, a POP or IMAP server is required as these protocols allow users to connect to their mailboxes from remote locations. IMAP offers several advantages over POP. These include the ability to store a copy of messages on a remote server after they are downloaded and concurrent updates. IMAP can be useful over low-speed links as it allows users to fetch the structure of messages without downloading them. It can also perform tasks such as searching on the server in order to minimize data transfer between clients and servers. Several POP and IMAP servers are available in the Ports Collection. These include mail/qpopper, mail/imap-uw, mail/courier-imap, and mail/dovecot2. It should be noted that both POP and IMAP transmit information, including username and password credentials, in clear-text. To secure the transmission of information across these protocols, consider tunneling sessions over &man.ssh.1; () or using SSL (). Domain Name System (DNS) The Domain Name System (DNS) and its daemon named play a large role in the delivery of email. In order to deliver mail from one site to another, the MTA will look up the remote site in DNS to determine which host will receive mail for the destination. This process also occurs when mail is sent from a remote host to the MTA. In addition to mapping hostnames to IP addresses, DNS is responsible for storing information specific to mail delivery, known as Mail eXchanger MX records. The MX record specifies which hosts will receive mail for a particular domain. To view the MX records for a domain, specify the type of record. Refer to &man.host.1;, for more details about this command: &prompt.user; host -t mx FreeBSD.org FreeBSD.org mail is handled by 10 mx1.FreeBSD.org Refer to for more information about DNS and its configuration. <application>Sendmail</application> Configuration Files Christopher Shumway Contributed by Sendmail Sendmail is the default MTA installed with &os;. It accepts mail from MUAs and delivers it to the appropriate mail host, as defined by its configuration. Sendmail can also accept network connections and deliver mail to local mailboxes or to another program. The configuration files for Sendmail are located in /etc/mail. This section describes these files in more detail. /etc/mail/access /etc/mail/aliases /etc/mail/local-host-names /etc/mail/mailer.conf /etc/mail/mailertable /etc/mail/sendmail.cf /etc/mail/virtusertable /etc/mail/access This access database file defines which hosts or IP addresses have access to the local mail server and what kind of access they have. Hosts listed as , which is the default option, are allowed to send mail to this host as long as the mail's final destination is the local machine. Hosts listed as are rejected for all mail connections. Hosts listed as are allowed to send mail for any destination using this mail server. Hosts listed as will have their mail returned with the specified mail error. If a host is listed as , Sendmail will abort the current search for this entry without accepting or rejecting the mail. Hosts listed as will have their messages held and will receive the specified text as the reason for the hold. Examples of using these options for both IPv4 and IPv6 addresses can be found in the &os; sample configuration, /etc/mail/access.sample: - # $FreeBSD$ + # $FreeBSD$ # # Mail relay access control list. Default is to reject mail unless the # destination is local, or listed in /etc/mail/local-host-names # ## Examples (commented out for safety) #From:cyberspammer.com ERROR:"550 We don't accept mail from spammers" #From:okay.cyberspammer.com OK #Connect:sendmail.org RELAY #To:sendmail.org RELAY #Connect:128.32 RELAY #Connect:128.32.2 SKIP #Connect:IPv6:1:2:3:4:5:6:7 RELAY #Connect:suspicious.example.com QUARANTINE:Mail from suspicious host #Connect:[127.0.0.3] OK #Connect:[IPv6:1:2:3:4:5:6:7:8] OK To configure the access database, use the format shown in the sample to make entries in /etc/mail/access, but do not put a comment symbol (#) in front of the entries. Create an entry for each host or network whose access should be configured. Mail senders that match the left side of the table are affected by the action on the right side of the table. Whenever this file is updated, update its database and restart Sendmail: &prompt.root; makemap hash /etc/mail/access < /etc/mail/access &prompt.root; service sendmail restart /etc/mail/aliases This database file contains a list of virtual mailboxes that are expanded to users, files, programs, or other aliases. Here are a few entries to illustrate the file format: root: localuser ftp-bugs: joe,eric,paul bit.bucket: /dev/null procmail: "|/usr/local/bin/procmail" The mailbox name on the left side of the colon is expanded to the target(s) on the right. The first entry expands the root mailbox to the localuser mailbox, which is then looked up in the /etc/mail/aliases database. If no match is found, the message is delivered to localuser. The second entry shows a mail list. Mail to ftp-bugs is expanded to the three local mailboxes joe, eric, and paul. A remote mailbox could be specified as user@example.com. The third entry shows how to write mail to a file, in this case /dev/null. The last entry demonstrates how to send mail to a program, /usr/local/bin/procmail, through a &unix; pipe. Refer to &man.aliases.5; for more information about the format of this file. Whenever this file is updated, run newaliases to update and initialize the aliases database. /etc/mail/sendmail.cf This is the master configuration file for Sendmail. It controls the overall behavior of Sendmail, including everything from rewriting email addresses to printing rejection messages to remote mail servers. Accordingly, this configuration file is quite complex. Fortunately, this file rarely needs to be changed for standard mail servers. The master Sendmail configuration file can be built from &man.m4.1; macros that define the features and behavior of Sendmail. Refer to /usr/src/contrib/sendmail/cf/README for some of the details. Whenever changes to this file are made, Sendmail needs to be restarted for the changes to take effect. /etc/mail/virtusertable This database file maps mail addresses for virtual domains and users to real mailboxes. These mailboxes can be local, remote, aliases defined in /etc/mail/aliases, or files. This allows multiple virtual domains to be hosted on one machine. &os; provides a sample configuration file in /etc/mail/virtusertable.sample to further demonstrate its format. The following example demonstrates how to create custom entries using that format: root@example.com root postmaster@example.com postmaster@noc.example.net @example.com joe This file is processed in a first match order. When an email address matches the address on the left, it is mapped to the local mailbox listed on the right. The format of the first entry in this example maps a specific email address to a local mailbox, whereas the format of the second entry maps a specific email address to a remote mailbox. Finally, any email address from example.com which has not matched any of the previous entries will match the last mapping and be sent to the local mailbox joe. When creating custom entries, use this format and add them to /etc/mail/virtusertable. Whenever this file is edited, update its database and restart Sendmail: &prompt.root; makemap hash /etc/mail/virtusertable < /etc/mail/virtusertable &prompt.root; service sendmail restart /etc/mail/relay-domains In a default &os; installation, Sendmail is configured to only send mail from the host it is running on. For example, if a POP server is available, users will be able to check mail from remote locations but they will not be able to send outgoing emails from outside locations. Typically, a few moments after the attempt, an email will be sent from MAILER-DAEMON with a 5.7 Relaying Denied message. The most straightforward solution is to add the ISP's FQDN to /etc/mail/relay-domains. If multiple addresses are needed, add them one per line: your.isp.example.com other.isp.example.net users-isp.example.org www.example.org After creating or editing this file, restart Sendmail with service sendmail restart. Now any mail sent through the system by any host in this list, provided the user has an account on the system, will succeed. This allows users to send mail from the system remotely without opening the system up to relaying SPAM from the Internet. Changing the Mail Transfer Agent Andrew Boothman Written by Gregory Neil Shapiro Information taken from emails written by email change mta &os; comes with Sendmail already installed as the MTA which is in charge of outgoing and incoming mail. However, the system administrator can change the system's MTA. A wide choice of alternative MTAs is available from the mail category of the &os; Ports Collection. Once a new MTA is installed, configure and test the new software before replacing Sendmail. Refer to the documentation of the new MTA for information on how to configure the software. Once the new MTA is working, use the instructions in this section to disable Sendmail and configure &os; to use the replacement MTA. Disable <application>Sendmail</application> If Sendmail's outgoing mail service is disabled, it is important that it is replaced with an alternative mail delivery system. Otherwise, system functions such as &man.periodic.8; will be unable to deliver their results by email. Many parts of the system expect a functional MTA. If applications continue to use Sendmail's binaries to try to send email after they are disabled, mail could go into an inactive Sendmail queue and never be delivered. In order to completely disable Sendmail, add or edit the following lines in /etc/rc.conf: sendmail_enable="NO" sendmail_submit_enable="NO" sendmail_outbound_enable="NO" sendmail_msp_queue_enable="NO" To only disable Sendmail's incoming mail service, use only this entry in /etc/rc.conf: sendmail_enable="NO" More information on Sendmail's startup options is available in &man.rc.sendmail.8;. Replace the Default <acronym>MTA</acronym> When a new MTA is installed using the Ports Collection, its startup script is also installed and startup instructions are mentioned in its package message. Before starting the new MTA, stop the running Sendmail processes. This example stops all of these services, then starts the Postfix service: &prompt.root; service sendmail stop &prompt.root; service postfix start To start the replacement MTA at system boot, add its configuration line to /etc/rc.conf. This entry enables the Postfix MTA: postfix_enable="YES" Some extra configuration is needed as Sendmail is so ubiquitous that some software assumes it is already installed and configured. Check /etc/periodic.conf and make sure that these values are set to NO. If this file does not exist, create it with these entries: daily_clean_hoststat_enable="NO" daily_status_mail_rejects_enable="NO" daily_status_include_submit_mailq="NO" daily_submit_queuerun="NO" Some alternative MTAs provide their own compatible implementations of the Sendmail command-line interface in order to facilitate using them as drop-in replacements for Sendmail. However, some MUAs may try to execute standard Sendmail binaries instead of the new MTA's binaries. &os; uses /etc/mail/mailer.conf to map the expected Sendmail binaries to the location of the new binaries. More information about this mapping can be found in &man.mailwrapper.8;. The default /etc/mail/mailer.conf looks like this: - # $FreeBSD$ + # $FreeBSD$ # # Execute the "real" sendmail program, named /usr/libexec/sendmail/sendmail # sendmail /usr/libexec/sendmail/sendmail send-mail /usr/libexec/sendmail/sendmail mailq /usr/libexec/sendmail/sendmail newaliases /usr/libexec/sendmail/sendmail hoststat /usr/libexec/sendmail/sendmail purgestat /usr/libexec/sendmail/sendmail When any of the commands listed on the left are run, the system actually executes the associated command shown on the right. This system makes it easy to change what binaries are executed when these default binaries are invoked. Some MTAs, when installed using the Ports Collection, will prompt to update this file for the new binaries. For example, Postfix will update the file like this: # # Execute the Postfix sendmail program, named /usr/local/sbin/sendmail # sendmail /usr/local/sbin/sendmail send-mail /usr/local/sbin/sendmail mailq /usr/local/sbin/sendmail newaliases /usr/local/sbin/sendmail If the installation of the MTA does not automatically update /etc/mail/mailer.conf, edit this file in a text editor so that it points to the new binaries. This example points to the binaries installed by mail/ssmtp: sendmail /usr/local/sbin/ssmtp send-mail /usr/local/sbin/ssmtp mailq /usr/local/sbin/ssmtp newaliases /usr/local/sbin/ssmtp hoststat /usr/bin/true purgestat /usr/bin/true Once everything is configured, it is recommended to reboot the system. Rebooting provides the opportunity to ensure that the system is correctly configured to start the new MTA automatically on boot. Troubleshooting email troubleshooting Why do I have to use the FQDN for hosts on my site? The host may actually be in a different domain. For example, in order for a host in foo.bar.edu to reach a host called mumble in the bar.edu domain, refer to it by the Fully-Qualified Domain Name FQDN, mumble.bar.edu, instead of just mumble. This is because the version of BIND BIND which ships with &os; no longer provides default abbreviations for non-FQDNs other than the local domain. An unqualified host such as mumble must either be found as mumble.foo.bar.edu, or it will be searched for in the root domain. In older versions of BIND, the search continued across mumble.bar.edu, and mumble.edu. RFC 1535 details why this is considered bad practice or even a security hole. As a good workaround, place the line: search foo.bar.edu bar.edu instead of the previous: domain foo.bar.edu into /etc/resolv.conf. However, make sure that the search order does not go beyond the boundary between local and public administration, as RFC 1535 calls it. How can I run a mail server on a dial-up PPP host? Connect to a &os; mail gateway on the LAN. The PPP connection is non-dedicated. One way to do this is to get a full-time Internet server to provide secondary MX MX record services for the domain. In this example, the domain is example.com and the ISP has configured example.net to provide secondary MX services to the domain: example.com. MX 10 example.com. MX 20 example.net. Only one host should be specified as the final recipient. For Sendmail, add Cw example.com in /etc/mail/sendmail.cf on example.com. When the sending MTA attempts to deliver mail, it will try to connect to the system, example.com, over the PPP link. This will time out if the destination is offline. The MTA will automatically deliver it to the secondary MX site at the Internet Service Provider (ISP), example.net. The secondary MX site will periodically try to connect to the primary MX host, example.com. Use something like this as a login script: #!/bin/sh # Put me in /usr/local/bin/pppmyisp ( sleep 60 ; /usr/sbin/sendmail -q ) & /usr/sbin/ppp -direct pppmyisp When creating a separate login script for users, instead use sendmail -qRexample.com in the script above. This will force all mail in the queue for example.com to be processed immediately. A further refinement of the situation can be seen from this example from the &a.isp;: > we provide the secondary MX for a customer. The customer connects to > our services several times a day automatically to get the mails to > his primary MX (We do not call his site when a mail for his domains > arrived). Our sendmail sends the mailqueue every 30 minutes. At the > moment he has to stay 30 minutes online to be sure that all mail is > gone to the primary MX. > > Is there a command that would initiate sendmail to send all the mails > now? The user has not root-privileges on our machine of course. In the privacy flags section of sendmail.cf, there is a definition Opgoaway,restrictqrun Remove restrictqrun to allow non-root users to start the queue processing. You might also like to rearrange the MXs. We are the 1st MX for our customers like this, and we have defined: # If we are the best MX for a host, try directly instead of generating # local config error. OwTrue That way a remote site will deliver straight to you, without trying the customer connection. You then send to your customer. Only works for hosts, so you need to get your customer to name their mail machine customer.com as well as hostname.customer.com in the DNS. Just put an A record in the DNS for customer.com. Advanced Topics This section covers more involved topics such as mail configuration and setting up mail for an entire domain. Basic Configuration email configuration Out of the box, one can send email to external hosts as long as /etc/resolv.conf is configured or the network has access to a configured DNS server. To have email delivered to the MTA on the &os; host, do one of the following: Run a DNS server for the domain. Get mail delivered directly to the FQDN for the machine. SMTP In order to have mail delivered directly to a host, it must have a permanent static IP address, not a dynamic IP address. If the system is behind a firewall, it must be configured to allow SMTP traffic. To receive mail directly at a host, one of these two must be configured: Make sure that the lowest-numbered MXMX record record in DNS points to the host's static IP address. Make sure there is no MX entry in the DNS for the host. Either of the above will allow mail to be received directly at the host. Try this: &prompt.root; hostname example.FreeBSD.org &prompt.root; host example.FreeBSD.org example.FreeBSD.org has address 204.216.27.XX In this example, mail sent directly to yourlogin@example.FreeBSD.org should work without problems, assuming Sendmail is running correctly on example.FreeBSD.org. For this example: &prompt.root; host example.FreeBSD.org example.FreeBSD.org has address 204.216.27.XX example.FreeBSD.org mail is handled (pri=10) by nevdull.FreeBSD.org All mail sent to example.FreeBSD.org will be collected on hub under the same username instead of being sent directly to your host. The above information is handled by the DNS server. The DNS record that carries mail routing information is the MX entry. If no MX record exists, mail will be delivered directly to the host by way of its IP address. The MX entry for freefall.FreeBSD.org at one time looked like this: freefall MX 30 mail.crl.net freefall MX 40 agora.rdrop.com freefall MX 10 freefall.FreeBSD.org freefall MX 20 who.cdrom.com freefall had many MX entries. The lowest MX number is the host that receives mail directly, if available. If it is not accessible for some reason, the next lower-numbered host will accept messages temporarily, and pass it along when a lower-numbered host becomes available. Alternate MX sites should have separate Internet connections in order to be most useful. Your ISP can provide this service. Mail for a Domain When configuring a MTA for a network, any mail sent to hosts in its domain should be diverted to the MTA so that users can receive their mail on the master mail server. DNS To make life easiest, a user account with the same username should exist on both the MTA and the system with the MUA. Use &man.adduser.8; to create the user accounts. The MTA must be the designated mail exchanger for each workstation on the network. This is done in theDNS configuration with an MX record: example.FreeBSD.org A 204.216.27.XX ; Workstation MX 10 nevdull.FreeBSD.org ; Mailhost This will redirect mail for the workstation to the MTA no matter where the A record points. The mail is sent to the MX host. This must be configured on a DNS server. If the network does not run its own DNS server, talk to the ISP or DNS provider. The following is an example of virtual email hosting. Consider a customer with the domain customer1.org, where all the mail for customer1.org should be sent to mail.myhost.com. The DNS entry should look like this: customer1.org MX 10 mail.myhost.com An A> record is not needed for customer1.org in order to only handle email for that domain. However, running ping against customer1.org will not work unless an A record exists for it. Tell the MTA which domains and/or hostnames it should accept mail for. Either of the following will work for Sendmail: Add the hosts to /etc/mail/local-host-names when using the FEATURE(use_cw_file). Add a Cwyour.host.com line to /etc/sendmail.cf. Setting Up to Send Only Bill Moran Contributed by There are many instances where one may only want to send mail through a relay. Some examples are: The computer is a desktop machine that needs to use programs such as &man.send-pr.1;, using the ISP's mail relay. The computer is a server that does not handle mail locally, but needs to pass off all mail to a relay for processing. While any MTA is capable of filling this particular niche, it can be difficult to properly configure a full-featured MTA just to handle offloading mail. Programs such as Sendmail and Postfix are overkill for this use. Additionally, a typical Internet access service agreement may forbid one from running a mail server. The easiest way to fulfill those needs is to install the mail/ssmtp port: &prompt.root; cd /usr/ports/mail/ssmtp &prompt.root; make install replace clean Once installed, mail/ssmtp can be configured with /usr/local/etc/ssmtp/ssmtp.conf: root=yourrealemail@example.com mailhub=mail.example.com rewriteDomain=example.com hostname=_HOSTNAME_ Use the real email address for root. Enter the ISP's outgoing mail relay in place of mail.example.com. Some ISPs call this the outgoing mail server or SMTP server. Make sure to disable Sendmail, including the outgoing mail service. See for details. mail/ssmtp has some other options available. Refer to the examples in /usr/local/etc/ssmtp or the manual page of ssmtp for more information. Setting up ssmtp in this manner allows any software on the computer that needs to send mail to function properly, while not violating the ISP's usage policy or allowing the computer to be hijacked for spamming. Using Mail with a Dialup Connection When using a static IP address, one should not need to adjust the default configuration. Set the hostname to the assigned Internet name and Sendmail will do the rest. When using a dynamically assigned IP address and a dialup PPP connection to the Internet, one usually has a mailbox on the ISP's mail server. In this example, the ISP's domain is example.net, the user name is user, the hostname is bsd.home, and the ISP has allowed relay.example.net as a mail relay. In order to retrieve mail from the ISP's mailbox, install a retrieval agent from the Ports Collection. mail/fetchmail is a good choice as it supports many different protocols. Usually, the ISP will provide POP. When using user PPP, email can be automatically fetched when an Internet connection is established with the following entry in /etc/ppp/ppp.linkup: MYADDR: !bg su user -c fetchmail When using Sendmail to deliver mail to non-local accounts, configure Sendmail to process the mail queue as soon as the Internet connection is established. To do this, add this line after the above fetchmail entry in /etc/ppp/ppp.linkup: !bg su user -c "sendmail -q" In this example, there is an account for user on bsd.home. In the home directory of user on bsd.home, create a .fetchmailrc which contains this line: poll example.net protocol pop3 fetchall pass MySecret This file should not be readable by anyone except user as it contains the password MySecret. In order to send mail with the correct from: header, configure Sendmail to use user@example.net rather than user@bsd.home and to send all mail via relay.example.net, allowing quicker mail transmission. The following .mc should suffice: VERSIONID(`bsd.home.mc version 1.0') OSTYPE(bsd4.4)dnl FEATURE(nouucp)dnl MAILER(local)dnl MAILER(smtp)dnl Cwlocalhost Cwbsd.home MASQUERADE_AS(`example.net')dnl FEATURE(allmasquerade)dnl FEATURE(masquerade_envelope)dnl FEATURE(nocanonify)dnl FEATURE(nodns)dnl define(`SMART_HOST', `relay.example.net') Dmbsd.home define(`confDOMAIN_NAME',`bsd.home')dnl define(`confDELIVERY_MODE',`deferred')dnl Refer to the previous section for details of how to convert this file into the sendmail.cf format. Do not forget to restart Sendmail after updating sendmail.cf. SMTP Authentication James Gorham Written by Configuring SMTP authentication on the MTA provides a number of benefits. SMTP authentication adds a layer of security to Sendmail, and provides mobile users who switch hosts the ability to use the same MTA without the need to reconfigure their mail client's settings each time. Install security/cyrus-sasl2 from the Ports Collection. This port supports a number of compile-time options. For the SMTP authentication method demonstrated in this example, make sure that is not disabled. After installing security/cyrus-sasl2, edit /usr/local/lib/sasl2/Sendmail.conf, or create it if it does not exist, and add the following line: pwcheck_method: saslauthd Next, install security/cyrus-sasl2-saslauthd and add the following line to /etc/rc.conf: saslauthd_enable="YES" Finally, start the saslauthd daemon: &prompt.root; service saslauthd start This daemon serves as a broker for Sendmail to authenticate against the &os; &man.passwd.5; database. This saves the trouble of creating a new set of usernames and passwords for each user that needs to use SMTP authentication, and keeps the login and mail password the same. Next, edit /etc/make.conf and add the following lines: SENDMAIL_CFLAGS=-I/usr/local/include/sasl -DSASL SENDMAIL_LDFLAGS=-L/usr/local/lib SENDMAIL_LDADD=-lsasl2 These lines provide Sendmail the proper configuration options for linking to cyrus-sasl2 at compile time. Make sure that cyrus-sasl2 has been installed before recompiling Sendmail. Recompile Sendmail by executing the following commands: &prompt.root; cd /usr/src/lib/libsmutil &prompt.root; make cleandir && make obj && make &prompt.root; cd /usr/src/lib/libsm &prompt.root; make cleandir && make obj && make &prompt.root; cd /usr/src/usr.sbin/sendmail &prompt.root; make cleandir && make obj && make && make install This compile should not have any problems if /usr/src has not changed extensively and the shared libraries it needs are available. After Sendmail has been compiled and reinstalled, edit /etc/mail/freebsd.mc or the local .mc. Many administrators choose to use the output from &man.hostname.1; as the name of .mc for uniqueness. Add these lines: dnl set SASL options TRUST_AUTH_MECH(`GSSAPI DIGEST-MD5 CRAM-MD5 LOGIN')dnl define(`confAUTH_MECHANISMS', `GSSAPI DIGEST-MD5 CRAM-MD5 LOGIN')dnl These options configure the different methods available to Sendmail for authenticating users. To use a method other than pwcheck, refer to the Sendmail documentation. Finally, run &man.make.1; while in /etc/mail. That will run the new .mc and create a .cf named either freebsd.cf or the name used for the local .mc. Then, run make install restart, which will copy the file to sendmail.cf, and properly restart Sendmail. For more information about this process, refer to /etc/mail/Makefile. To test the configuration, use a MUA to send a test message. For further investigation, set the of Sendmail to 13 and watch /var/log/maillog for any errors. For more information, refer to SMTP authentication. Mail User Agents Marc Silver Contributed by Mail User Agents A MUA is an application that is used to send and receive email. As email evolves and becomes more complex, MUAs are becoming increasingly powerful and provide users increased functionality and flexibility. The mail category of the &os; Ports Collection contains numerous MUAs. These include graphical email clients such as Evolution or Balsa and console based clients such as mutt or alpine. <command>mail</command> &man.mail.1; is the default MUA installed with &os;. It is a console based MUA that offers the basic functionality required to send and receive text-based email. It provides limited attachment support and can only access local mailboxes. Although mail does not natively support interaction with POP or IMAP servers, these mailboxes may be downloaded to a local mbox using an application such as fetchmail. In order to send and receive email, run mail: &prompt.user; mail The contents of the user's mailbox in /var/mail are automatically read by mail. Should the mailbox be empty, the utility exits with a message indicating that no mail could be found. If mail exists, the application interface starts, and a list of messages will be displayed. Messages are automatically numbered, as can be seen in the following example: Mail version 8.1 6/6/93. Type ? for help. "/var/mail/marcs": 3 messages 3 new >N 1 root@localhost Mon Mar 8 14:05 14/510 "test" N 2 root@localhost Mon Mar 8 14:05 14/509 "user account" N 3 root@localhost Mon Mar 8 14:05 14/509 "sample" Messages can now be read by typing t followed by the message number. This example reads the first email: & t 1 Message 1: From root@localhost Mon Mar 8 14:05:52 2004 X-Original-To: marcs@localhost Delivered-To: marcs@localhost To: marcs@localhost Subject: test Date: Mon, 8 Mar 2004 14:05:52 +0200 (SAST) From: root@localhost (Charlie Root) This is a test message, please reply if you receive it. As seen in this example, the message will be displayed with full headers. To display the list of messages again, press h. If the email requires a reply, press either R or r mail keys. R instructs mail to reply only to the sender of the email, while r replies to all other recipients of the message. These commands can be suffixed with the mail number of the message to reply to. After typing the response, the end of the message should be marked by a single . on its own line. An example can be seen below: & R 1 To: root@localhost Subject: Re: test Thank you, I did get your email. . EOT In order to send a new email, press m, followed by the recipient email address. Multiple recipients may be specified by separating each address with the , delimiter. The subject of the message may then be entered, followed by the message contents. The end of the message should be specified by putting a single . on its own line. & mail root@localhost Subject: I mastered mail Now I can send and receive email using mail ... :) . EOT While using mail, press ? to display help at any time. Refer to &man.mail.1; for more help on how to use mail. &man.mail.1; was not designed to handle attachments and thus deals with them poorly. Newer MUAs handle attachments in a more intelligent way. Users who prefer to use mail may find the converters/mpack port to be of considerable use. <application>mutt</application> mutt is a powerful MUA, with many features, including: The ability to thread messages. PGP support for digital signing and encryption of email. MIME support. Maildir support. Highly customizable. Refer to http://www.mutt.org for more information on mutt. mutt may be installed using the mail/mutt port. After the port has been installed, mutt can be started by issuing the following command: &prompt.user; mutt mutt will automatically read and display the contents of the user mailbox in /var/mail. If no mails are found, mutt will wait for commands from the user. The example below shows mutt displaying a list of messages: To read an email, select it using the cursor keys and press Enter. An example of mutt displaying email can be seen below: Similar to &man.mail.1;, mutt can be used to reply only to the sender of the message as well as to all recipients. To reply only to the sender of the email, press r. To send a group reply to the original sender as well as all the message recipients, press g. By default, mutt uses the &man.vi.1; editor for creating and replying to emails. Each user can customize this by creating or editing the .muttrc in their home directory and setting the editor variable or by setting the EDITOR environment variable. Refer to http://www.mutt.org/ for more information about configuring mutt. To compose a new mail message, press m. After a valid subject has been given, mutt will start &man.vi.1; so the email can be written. Once the contents of the email are complete, save and quit from vi. mutt will resume, displaying a summary screen of the mail that is to be delivered. In order to send the mail, press y. An example of the summary screen can be seen below: mutt contains extensive help which can be accessed from most of the menus by pressing ?. The top line also displays the keyboard shortcuts where appropriate. <application>alpine</application> alpine is aimed at a beginner user, but also includes some advanced features. alpine has had several remote vulnerabilities discovered in the past, which allowed remote attackers to execute arbitrary code as users on the local system, by the action of sending a specially-prepared email. While known problems have been fixed, alpine code is written in an insecure style and the &os; Security Officer believes there are likely to be other undiscovered vulnerabilities. Users install alpine at their own risk. The current version of alpine may be installed using the mail/alpine port. Once the port has installed, alpine can be started by issuing the following command: &prompt.user; alpine The first time alpine runs, it displays a greeting page with a brief introduction, as well as a request from the alpine development team to send an anonymous email message allowing them to judge how many users are using their client. To send this anonymous message, press Enter. Alternatively, press E to exit the greeting without sending an anonymous message. An example of the greeting page is shown below: The main menu is then presented, which can be navigated using the cursor keys. This main menu provides shortcuts for the composing new mails, browsing mail directories, and administering address book entries. Below the main menu, relevant keyboard shortcuts to perform functions specific to the task at hand are shown. The default directory opened by alpine is inbox. To view the message index, press I, or select the MESSAGE INDEX option shown below: The message index shows messages in the current directory and can be navigated by using the cursor keys. Highlighted messages can be read by pressing Enter. In the screenshot below, a sample message is displayed by alpine. Contextual keyboard shortcuts are displayed at the bottom of the screen. An example of one of a shortcut is r, which tells the MUA to reply to the current message being displayed. Replying to an email in alpine is done using the pico editor, which is installed by default with alpine. pico makes it easy to navigate the message and is easier for novice users to use than &man.vi.1; or &man.mail.1;. Once the reply is complete, the message can be sent by pressing CtrlX . alpine will ask for confirmation before sending the message. alpine can be customized using the SETUP option from the main menu. Consult http://www.washington.edu/alpine/ for more information. Using <application>fetchmail</application> Marc Silver Contributed by fetchmail fetchmail is a full-featured IMAP and POP client. It allows users to automatically download mail from remote IMAP and POP servers and save it into local mailboxes where it can be accessed more easily. fetchmail can be installed using the mail/fetchmail port, and offers various features, including: Support for the POP3, APOP, KPOP, IMAP, ETRN and ODMR protocols. Ability to forward mail using SMTP, which allows filtering, forwarding, and aliasing to function normally. May be run in daemon mode to check periodically for new messages. Can retrieve multiple mailboxes and forward them, based on configuration, to different local users. This section explains some of the basic features of fetchmail. This utility requires a .fetchmailrc configuration in the user's home directory in order to run correctly. This file includes server information as well as login credentials. Due to the sensitive nature of the contents of this file, it is advisable to make it readable only by the user, with the following command: &prompt.user; chmod 600 .fetchmailrc The following .fetchmailrc serves as an example for downloading a single user mailbox using POP. It tells fetchmail to connect to example.com using a username of joesoap and a password of XXX. This example assumes that the user joesoap exists on the local system. poll example.com protocol pop3 username "joesoap" password "XXX" The next example connects to multiple POP and IMAP servers and redirects to different local usernames where applicable: poll example.com proto pop3: user "joesoap", with password "XXX", is "jsoap" here; user "andrea", with password "XXXX"; poll example2.net proto imap: user "john", with password "XXXXX", is "myth" here; fetchmail can be run in daemon mode by running it with , followed by the interval (in seconds) that fetchmail should poll servers listed in .fetchmailrc. The following example configures fetchmail to poll every 600 seconds: &prompt.user; fetchmail -d 600 More information on fetchmail can be found at http://www.fetchmail.info/. Using <application>procmail</application> Marc Silver Contributed by procmail procmail is a powerful application used to filter incoming mail. It allows users to define rules which can be matched to incoming mails to perform specific functions or to reroute mail to alternative mailboxes or email addresses. procmail can be installed using the mail/procmail port. Once installed, it can be directly integrated into most MTAs. Consult the MTA documentation for more information. Alternatively, procmail can be integrated by adding the following line to a .forward in the home directory of the user: "|exec /usr/local/bin/procmail || exit 75" The following section displays some basic procmail rules, as well as brief descriptions of what they do. Rules must be inserted into a .procmailrc, which must reside in the user's home directory. The majority of these rules can be found in &man.procmailex.5;. To forward all mail from user@example.com to an external address of goodmail@example2.com: :0 * ^From.*user@example.com ! goodmail@example2.com To forward all mails shorter than 1000 bytes to an external address of goodmail@example2.com: :0 * < 1000 ! goodmail@example2.com To send all mail sent to alternate@example.com to a mailbox called alternate: :0 * ^TOalternate@example.com alternate To send all mail with a subject of Spam to /dev/null: :0 ^Subject:.*Spam /dev/null A useful recipe that parses incoming &os;.org mailing lists and places each list in its own mailbox: :0 * ^Sender:.owner-freebsd-\/[^@]+@FreeBSD.ORG { LISTNAME=${MATCH} :0 * LISTNAME??^\/[^@]+ FreeBSD-${MATCH} } Index: head/en_US.ISO8859-1/books/handbook/network-servers/chapter.xml =================================================================== --- head/en_US.ISO8859-1/books/handbook/network-servers/chapter.xml (revision 50807) +++ head/en_US.ISO8859-1/books/handbook/network-servers/chapter.xml (revision 50808) @@ -1,5786 +1,5786 @@ Network Servers Synopsis This chapter covers some of the more frequently used network services on &unix; systems. This includes installing, configuring, testing, and maintaining many different types of network services. Example configuration files are included throughout this chapter for reference. By the end of this chapter, readers will know: How to manage the inetd daemon. How to set up the Network File System (NFS). How to set up the Network Information Server (NIS) for centralizing and sharing user accounts. How to set &os; up to act as an LDAP server or client How to set up automatic network settings using DHCP. How to set up a Domain Name Server (DNS). How to set up the Apache HTTP Server. How to set up a File Transfer Protocol (FTP) server. How to set up a file and print server for &windows; clients using Samba. How to synchronize the time and date, and set up a time server using the Network Time Protocol (NTP). How to set up iSCSI. This chapter assumes a basic knowledge of: /etc/rc scripts. Network terminology. Installation of additional third-party software (). The <application>inetd</application> Super-Server The &man.inetd.8; daemon is sometimes referred to as a Super-Server because it manages connections for many services. Instead of starting multiple applications, only the inetd service needs to be started. When a connection is received for a service that is managed by inetd, it determines which program the connection is destined for, spawns a process for that program, and delegates the program a socket. Using inetd for services that are not heavily used can reduce system load, when compared to running each daemon individually in stand-alone mode. Primarily, inetd is used to spawn other daemons, but several trivial protocols are handled internally, such as chargen, auth, time, echo, discard, and daytime. This section covers the basics of configuring inetd. Configuration File Configuration of inetd is done by editing /etc/inetd.conf. Each line of this configuration file represents an application which can be started by inetd. By default, every line starts with a comment (#), meaning that inetd is not listening for any applications. To configure inetd to listen for an application's connections, remove the # at the beginning of the line for that application. After saving your edits, configure inetd to start at system boot by editing /etc/rc.conf: inetd_enable="YES" To start inetd now, so that it listens for the service you configured, type: &prompt.root; service inetd start Once inetd is started, it needs to be notified whenever a modification is made to /etc/inetd.conf: Reloading the <application>inetd</application> Configuration File &prompt.root; service inetd reload Typically, the default entry for an application does not need to be edited beyond removing the #. In some situations, it may be appropriate to edit the default entry. As an example, this is the default entry for &man.ftpd.8; over IPv4: ftp stream tcp nowait root /usr/libexec/ftpd ftpd -l The seven columns in an entry are as follows: service-name socket-type protocol {wait|nowait}[/max-child[/max-connections-per-ip-per-minute[/max-child-per-ip]]] user[:group][/login-class] server-program server-program-arguments where: service-name The service name of the daemon to start. It must correspond to a service listed in /etc/services. This determines which port inetd listens on for incoming connections to that service. When using a custom service, it must first be added to /etc/services. socket-type Either stream, dgram, raw, or seqpacket. Use stream for TCP connections and dgram for UDP services. protocol Use one of the following protocol names: Protocol Name Explanation tcp or tcp4 TCP IPv4 udp or udp4 UDP IPv4 tcp6 TCP IPv6 udp6 UDP IPv6 tcp46 Both TCP IPv4 and IPv6 udp46 Both UDP IPv4 and IPv6 {wait|nowait}[/max-child[/max-connections-per-ip-per-minute[/max-child-per-ip]]] In this field, or must be specified. , and are optional. indicates whether or not the service is able to handle its own socket. socket types must use while daemons, which are usually multi-threaded, should use . usually hands off multiple sockets to a single daemon, while spawns a child daemon for each new socket. The maximum number of child daemons inetd may spawn is set by . For example, to limit ten instances of the daemon, place a /10 after . Specifying /0 allows an unlimited number of children. limits the number of connections from any particular IP address per minute. Once the limit is reached, further connections from this IP address will be dropped until the end of the minute. For example, a value of /10 would limit any particular IP address to ten connection attempts per minute. limits the number of child processes that can be started on behalf on any single IP address at any moment. These options can limit excessive resource consumption and help to prevent Denial of Service attacks. An example can be seen in the default settings for &man.fingerd.8;: finger stream tcp nowait/3/10 nobody /usr/libexec/fingerd fingerd -k -s user The username the daemon will run as. Daemons typically run as root, daemon, or nobody. server-program The full path to the daemon. If the daemon is a service provided by inetd internally, use . server-program-arguments Used to specify any command arguments to be passed to the daemon on invocation. If the daemon is an internal service, use . Command-Line Options Like most server daemons, inetd has a number of options that can be used to modify its behavior. By default, inetd is started with -wW -C 60. These options enable TCP wrappers for all services, including internal services, and prevent any IP address from requesting any service more than 60 times per minute. To change the default options which are passed to inetd, add an entry for inetd_flags in /etc/rc.conf. If inetd is already running, restart it with service inetd restart. The available rate limiting options are: -c maximum Specify the default maximum number of simultaneous invocations of each service, where the default is unlimited. May be overridden on a per-service basis by using in /etc/inetd.conf. -C rate Specify the default maximum number of times a service can be invoked from a single IP address per minute. May be overridden on a per-service basis by using in /etc/inetd.conf. -R rate Specify the maximum number of times a service can be invoked in one minute, where the default is 256. A rate of 0 allows an unlimited number. -s maximum Specify the maximum number of times a service can be invoked from a single IP address at any one time, where the default is unlimited. May be overridden on a per-service basis by using in /etc/inetd.conf. Additional options are available. Refer to &man.inetd.8; for the full list of options. Security Considerations Many of the daemons which can be managed by inetd are not security-conscious. Some daemons, such as fingerd, can provide information that may be useful to an attacker. Only enable the services which are needed and monitor the system for excessive connection attempts. max-connections-per-ip-per-minute, max-child and max-child-per-ip can be used to limit such attacks. By default, TCP wrappers is enabled. Consult &man.hosts.access.5; for more information on placing TCP restrictions on various inetd invoked daemons. Network File System (NFS) Tom Rhodes Reorganized and enhanced by Bill Swingle Written by NFS &os; supports the Network File System (NFS), which allows a server to share directories and files with clients over a network. With NFS, users and programs can access files on remote systems as if they were stored locally. NFS has many practical uses. Some of the more common uses include: Data that would otherwise be duplicated on each client can be kept in a single location and accessed by clients on the network. Several clients may need access to the /usr/ports/distfiles directory. Sharing that directory allows for quick access to the source files without having to download them to each client. On large networks, it is often more convenient to configure a central NFS server on which all user home directories are stored. Users can log into a client anywhere on the network and have access to their home directories. Administration of NFS exports is simplified. For example, there is only one file system where security or backup policies must be set. Removable media storage devices can be used by other machines on the network. This reduces the number of devices throughout the network and provides a centralized location to manage their security. It is often more convenient to install software on multiple machines from a centralized installation media. NFS consists of a server and one or more clients. The client remotely accesses the data that is stored on the server machine. In order for this to function properly, a few processes have to be configured and running. These daemons must be running on the server: NFS server file server UNIX clients rpcbind mountd nfsd Daemon Description nfsd The NFS daemon which services requests from NFS clients. mountd The NFS mount daemon which carries out requests received from nfsd. rpcbind This daemon allows NFS clients to discover which port the NFS server is using. Running &man.nfsiod.8; on the client can improve performance, but is not required. Configuring the Server NFS configuration The file systems which the NFS server will share are specified in /etc/exports. Each line in this file specifies a file system to be exported, which clients have access to that file system, and any access options. When adding entries to this file, each exported file system, its properties, and allowed hosts must occur on a single line. If no clients are listed in the entry, then any client on the network can mount that file system. NFS export examples The following /etc/exports entries demonstrate how to export file systems. The examples can be modified to match the file systems and client names on the reader's network. There are many options that can be used in this file, but only a few will be mentioned here. See &man.exports.5; for the full list of options. This example shows how to export /cdrom to three hosts named alpha, bravo, and charlie: /cdrom -ro alpha bravo charlie The -ro flag makes the file system read-only, preventing clients from making any changes to the exported file system. This example assumes that the host names are either in DNS or in /etc/hosts. Refer to &man.hosts.5; if the network does not have a DNS server. The next example exports /home to three clients by IP address. This can be useful for networks without DNS or /etc/hosts entries. The -alldirs flag allows subdirectories to be mount points. In other words, it will not automatically mount the subdirectories, but will permit the client to mount the directories that are required as needed. /usr/home -alldirs 10.0.0.2 10.0.0.3 10.0.0.4 This next example exports /a so that two clients from different domains may access that file system. The allows root on the remote system to write data on the exported file system as root. If -maproot=root is not specified, the client's root user will be mapped to the server's nobody account and will be subject to the access limitations defined for nobody. /a -maproot=root host.example.com box.example.org A client can only be specified once per file system. For example, if /usr is a single file system, these entries would be invalid as both entries specify the same host: # Invalid when /usr is one file system /usr/src client /usr/ports client The correct format for this situation is to use one entry: /usr/src /usr/ports client The following is an example of a valid export list, where /usr and /exports are local file systems: # Export src and ports to client01 and client02, but only # client01 has root privileges on it /usr/src /usr/ports -maproot=root client01 /usr/src /usr/ports client02 # The client machines have root and can mount anywhere # on /exports. Anyone in the world can mount /exports/obj read-only /exports -alldirs -maproot=root client01 client02 /exports/obj -ro To enable the processes required by the NFS server at boot time, add these options to /etc/rc.conf: rpcbind_enable="YES" nfs_server_enable="YES" mountd_flags="-r" The server can be started now by running this command: &prompt.root; service nfsd start Whenever the NFS server is started, mountd also starts automatically. However, mountd only reads /etc/exports when it is started. To make subsequent /etc/exports edits take effect immediately, force mountd to reread it: &prompt.root; service mountd reload Configuring the Client To enable NFS clients, set this option in each client's /etc/rc.conf: nfs_client_enable="YES" Then, run this command on each NFS client: &prompt.root; service nfsclient start The client now has everything it needs to mount a remote file system. In these examples, the server's name is server and the client's name is client. To mount /home on server to the /mnt mount point on client: NFS mounting &prompt.root; mount server:/home /mnt The files and directories in /home will now be available on client, in the /mnt directory. To mount a remote file system each time the client boots, add it to /etc/fstab: server:/home /mnt nfs rw 0 0 Refer to &man.fstab.5; for a description of all available options. Locking Some applications require file locking to operate correctly. To enable locking, add these lines to /etc/rc.conf on both the client and server: rpc_lockd_enable="YES" rpc_statd_enable="YES" Then start the applications: &prompt.root; service lockd start &prompt.root; service statd start If locking is not required on the server, the NFS client can be configured to lock locally by including when running mount. Refer to &man.mount.nfs.8; for further details. Automating Mounts with &man.amd.8; Wylie Stilwell Contributed by Chern Lee Rewritten by amd automatic mounter daemon The automatic mounter daemon, amd, automatically mounts a remote file system whenever a file or directory within that file system is accessed. File systems that are inactive for a period of time will be automatically unmounted by amd. This daemon provides an alternative to modifying /etc/fstab to list every client. It operates by attaching itself as an NFS server to the /host and /net directories. When a file is accessed within one of these directories, amd looks up the corresponding remote mount and automatically mounts it. /net is used to mount an exported file system from an IP address while /host is used to mount an export from a remote hostname. For instance, an attempt to access a file within /host/foobar/usr would tell amd to mount the /usr export on the host foobar. Mounting an Export with <application>amd</application> In this example, showmount -e shows the exported file systems that can be mounted from the NFS server, foobar: &prompt.user; showmount -e foobar Exports list on foobar: /usr 10.10.10.0 /a 10.10.10.0 &prompt.user; cd /host/foobar/usr The output from showmount shows /usr as an export. When changing directories to /host/foobar/usr, amd intercepts the request and attempts to resolve the hostname foobar. If successful, amd automatically mounts the desired export. To enable amd at boot time, add this line to /etc/rc.conf: amd_enable="YES" To start amd now: &prompt.root; service amd start Custom flags can be passed to amd from the amd_flags environment variable. By default, amd_flags is set to: amd_flags="-a /.amd_mnt -l syslog /host /etc/amd.map /net /etc/amd.map" The default options with which exports are mounted are defined in /etc/amd.map. Some of the more advanced features of amd are defined in /etc/amd.conf. Consult &man.amd.8; and &man.amd.conf.5; for more information. Automating Mounts with &man.autofs.5; The &man.autofs.5; automount facility is supported starting with &os; 10.1-RELEASE. To use the automounter functionality in older versions of &os;, use &man.amd.8; instead. This chapter only describes the &man.autofs.5; automounter. autofs automounter subsystem The &man.autofs.5; facility is a common name for several components that, together, allow for automatic mounting of remote and local filesystems whenever a file or directory within that file system is accessed. It consists of the kernel component, &man.autofs.5;, and several userspace applications: &man.automount.8;, &man.automountd.8; and &man.autounmountd.8;. It serves as an alternative for &man.amd.8; from previous &os; releases. Amd is still provided for backward compatibility purposes, as the two use different map format; the one used by autofs is the same as with other SVR4 automounters, such as the ones in Solaris, MacOS X, and Linux. The &man.autofs.5; virtual filesystem is mounted on specified mountpoints by &man.automount.8;, usually invoked during boot. Whenever a process attempts to access file within the &man.autofs.5; mountpoint, the kernel will notify &man.automountd.8; daemon and pause the triggering process. The &man.automountd.8; daemon will handle kernel requests by finding the proper map and mounting the filesystem according to it, then signal the kernel to release blocked process. The &man.autounmountd.8; daemon automatically unmounts automounted filesystems after some time, unless they are still being used. The primary autofs configuration file is /etc/auto_master. It assigns individual maps to top-level mounts. For an explanation of auto_master and the map syntax, refer to &man.auto.master.5;. There is a special automounter map mounted on /net. When a file is accessed within this directory, &man.autofs.5; looks up the corresponding remote mount and automatically mounts it. For instance, an attempt to access a file within /net/foobar/usr would tell &man.automountd.8; to mount the /usr export from the host foobar. Mounting an Export with &man.autofs.5; In this example, showmount -e shows the exported file systems that can be mounted from the NFS server, foobar: &prompt.user; showmount -e foobar Exports list on foobar: /usr 10.10.10.0 /a 10.10.10.0 &prompt.user; cd /net/foobar/usr The output from showmount shows /usr as an export. When changing directories to /host/foobar/usr, &man.automountd.8; intercepts the request and attempts to resolve the hostname foobar. If successful, &man.automountd.8; automatically mounts the source export. To enable &man.autofs.5; at boot time, add this line to /etc/rc.conf: autofs_enable="YES" Then &man.autofs.5; can be started by running: &prompt.root; service automount start &prompt.root; service automountd start &prompt.root; service autounmountd start The &man.autofs.5; map format is the same as in other operating systems. Information about this format from other sources can be useful, like the Mac OS X document. Consult the &man.automount.8;, &man.automountd.8;, &man.autounmountd.8;, and &man.auto.master.5; manual pages for more information. Network Information System (<acronym>NIS</acronym>) NIS Solaris HP-UX AIX Linux NetBSD OpenBSD yellow pages NIS Network Information System (NIS) is designed to centralize administration of &unix;-like systems such as &solaris;, HP-UX, &aix;, Linux, NetBSD, OpenBSD, and &os;. NIS was originally known as Yellow Pages but the name was changed due to trademark issues. This is the reason why NIS commands begin with yp. NIS domains NIS is a Remote Procedure Call (RPC)-based client/server system that allows a group of machines within an NIS domain to share a common set of configuration files. This permits a system administrator to set up NIS client systems with only minimal configuration data and to add, remove, or modify configuration data from a single location. &os; uses version 2 of the NIS protocol. <acronym>NIS</acronym> Terms and Processes Table 28.1 summarizes the terms and important processes used by NIS: rpcbind portmap <acronym>NIS</acronym> Terminology Term Description NIS domain name NIS servers and clients share an NIS domain name. Typically, this name does not have anything to do with DNS. &man.rpcbind.8; This service enables RPC and must be running in order to run an NIS server or act as an NIS client. &man.ypbind.8; This service binds an NIS client to its NIS server. It will take the NIS domain name and use RPC to connect to the server. It is the core of client/server communication in an NIS environment. If this service is not running on a client machine, it will not be able to access the NIS server. &man.ypserv.8; This is the process for the NIS server. If this service stops running, the server will no longer be able to respond to NIS requests so hopefully, there is a slave server to take over. Some non-&os; clients will not try to reconnect using a slave server and the ypbind process may need to be restarted on these clients. &man.rpc.yppasswdd.8; This process only runs on NIS master servers. This daemon allows NIS clients to change their NIS passwords. If this daemon is not running, users will have to login to the NIS master server and change their passwords there.
Machine Types NIS master server NIS slave server NIS client There are three types of hosts in an NIS environment: NIS master server This server acts as a central repository for host configuration information and maintains the authoritative copy of the files used by all of the NIS clients. The passwd, group, and other various files used by NIS clients are stored on the master server. While it is possible for one machine to be an NIS master server for more than one NIS domain, this type of configuration will not be covered in this chapter as it assumes a relatively small-scale NIS environment. NIS slave servers NIS slave servers maintain copies of the NIS master's data files in order to provide redundancy. Slave servers also help to balance the load of the master server as NIS clients always attach to the NIS server which responds first. NIS clients NIS clients authenticate against the NIS server during log on. Information in many files can be shared using NIS. The master.passwd, group, and hosts files are commonly shared via NIS. Whenever a process on a client needs information that would normally be found in these files locally, it makes a query to the NIS server that it is bound to instead. Planning Considerations This section describes a sample NIS environment which consists of 15 &os; machines with no centralized point of administration. Each machine has its own /etc/passwd and /etc/master.passwd. These files are kept in sync with each other only through manual intervention. Currently, when a user is added to the lab, the process must be repeated on all 15 machines. The configuration of the lab will be as follows: Machine name IP address Machine role ellington 10.0.0.2 NIS master coltrane 10.0.0.3 NIS slave basie 10.0.0.4 Faculty workstation bird 10.0.0.5 Client machine cli[1-11] 10.0.0.[6-17] Other client machines If this is the first time an NIS scheme is being developed, it should be thoroughly planned ahead of time. Regardless of network size, several decisions need to be made as part of the planning process. Choosing a <acronym>NIS</acronym> Domain Name NIS domain name When a client broadcasts its requests for info, it includes the name of the NIS domain that it is part of. This is how multiple servers on one network can tell which server should answer which request. Think of the NIS domain name as the name for a group of hosts. Some organizations choose to use their Internet domain name for their NIS domain name. This is not recommended as it can cause confusion when trying to debug network problems. The NIS domain name should be unique within the network and it is helpful if it describes the group of machines it represents. For example, the Art department at Acme Inc. might be in the acme-art NIS domain. This example will use the domain name test-domain. However, some non-&os; operating systems require the NIS domain name to be the same as the Internet domain name. If one or more machines on the network have this restriction, the Internet domain name must be used as the NIS domain name. Physical Server Requirements There are several things to keep in mind when choosing a machine to use as a NIS server. Since NIS clients depend upon the availability of the server, choose a machine that is not rebooted frequently. The NIS server should ideally be a stand alone machine whose sole purpose is to be an NIS server. If the network is not heavily used, it is acceptable to put the NIS server on a machine running other services. However, if the NIS server becomes unavailable, it will adversely affect all NIS clients. Configuring the <acronym>NIS</acronym> Master Server The canonical copies of all NIS files are stored on the master server. The databases used to store the information are called NIS maps. In &os;, these maps are stored in /var/yp/[domainname] where [domainname] is the name of the NIS domain. Since multiple domains are supported, it is possible to have several directories, one for each domain. Each domain will have its own independent set of maps. NIS master and slave servers handle all NIS requests through &man.ypserv.8;. This daemon is responsible for receiving incoming requests from NIS clients, translating the requested domain and map name to a path to the corresponding database file, and transmitting data from the database back to the client. NIS server configuration Setting up a master NIS server can be relatively straight forward, depending on environmental needs. Since &os; provides built-in NIS support, it only needs to be enabled by adding the following lines to /etc/rc.conf: nisdomainname="test-domain" nis_server_enable="YES" nis_yppasswdd_enable="YES" This line sets the NIS domain name to test-domain. This automates the start up of the NIS server processes when the system boots. This enables the &man.rpc.yppasswdd.8; daemon so that users can change their NIS password from a client machine. Care must be taken in a multi-server domain where the server machines are also NIS clients. It is generally a good idea to force the servers to bind to themselves rather than allowing them to broadcast bind requests and possibly become bound to each other. Strange failure modes can result if one server goes down and others are dependent upon it. Eventually, all the clients will time out and attempt to bind to other servers, but the delay involved can be considerable and the failure mode is still present since the servers might bind to each other all over again. A server that is also a client can be forced to bind to a particular server by adding these additional lines to /etc/rc.conf: nis_client_enable="YES" # run client stuff as well nis_client_flags="-S NIS domain,server" After saving the edits, type /etc/netstart to restart the network and apply the values defined in /etc/rc.conf. Before initializing the NIS maps, start &man.ypserv.8;: &prompt.root; service ypserv start Initializing the <acronym>NIS</acronym> Maps NIS maps NIS maps are generated from the configuration files in /etc on the NIS master, with one exception: /etc/master.passwd. This is to prevent the propagation of passwords to all the servers in the NIS domain. Therefore, before the NIS maps are initialized, configure the primary password files: &prompt.root; cp /etc/master.passwd /var/yp/master.passwd &prompt.root; cd /var/yp &prompt.root; vi master.passwd It is advisable to remove all entries for system accounts as well as any user accounts that do not need to be propagated to the NIS clients, such as the root and any other administrative accounts. Ensure that the /var/yp/master.passwd is neither group or world readable by setting its permissions to 600. After completing this task, initialize the NIS maps. &os; includes the &man.ypinit.8; script to do this. When generating maps for the master server, include and specify the NIS domain name: ellington&prompt.root; ypinit -m test-domain Server Type: MASTER Domain: test-domain Creating an YP server will require that you answer a few questions. Questions will all be asked at the beginning of the procedure. Do you want this procedure to quit on non-fatal errors? [y/n: n] n Ok, please remember to go back and redo manually whatever fails. If not, something might not work. At this point, we have to construct a list of this domains YP servers. rod.darktech.org is already known as master server. Please continue to add any slave servers, one per line. When you are done with the list, type a <control D>. master server : ellington next host to add: coltrane next host to add: ^D The current list of NIS servers looks like this: ellington coltrane Is this correct? [y/n: y] y [..output from map generation..] NIS Map update completed. ellington has been setup as an YP master server without any errors. This will create /var/yp/Makefile from /var/yp/Makefile.dist. By default, this file assumes that the environment has a single NIS server with only &os; clients. Since test-domain has a slave server, edit this line in /var/yp/Makefile so that it begins with a comment (#): NOPUSH = "True" Adding New Users Every time a new user is created, the user account must be added to the master NIS server and the NIS maps rebuilt. Until this occurs, the new user will not be able to login anywhere except on the NIS master. For example, to add the new user jsmith to the test-domain domain, run these commands on the master server: &prompt.root; pw useradd jsmith &prompt.root; cd /var/yp &prompt.root; make test-domain The user could also be added using adduser jsmith instead of pw useradd smith. Setting up a <acronym>NIS</acronym> Slave Server NIS slave server To set up an NIS slave server, log on to the slave server and edit /etc/rc.conf as for the master server. Do not generate any NIS maps, as these already exist on the master server. When running ypinit on the slave server, use (for slave) instead of (for master). This option requires the name of the NIS master in addition to the domain name, as seen in this example: coltrane&prompt.root; ypinit -s ellington test-domain Server Type: SLAVE Domain: test-domain Master: ellington Creating an YP server will require that you answer a few questions. Questions will all be asked at the beginning of the procedure. Do you want this procedure to quit on non-fatal errors? [y/n: n] n Ok, please remember to go back and redo manually whatever fails. If not, something might not work. There will be no further questions. The remainder of the procedure should take a few minutes, to copy the databases from ellington. Transferring netgroup... ypxfr: Exiting: Map successfully transferred Transferring netgroup.byuser... ypxfr: Exiting: Map successfully transferred Transferring netgroup.byhost... ypxfr: Exiting: Map successfully transferred Transferring master.passwd.byuid... ypxfr: Exiting: Map successfully transferred Transferring passwd.byuid... ypxfr: Exiting: Map successfully transferred Transferring passwd.byname... ypxfr: Exiting: Map successfully transferred Transferring group.bygid... ypxfr: Exiting: Map successfully transferred Transferring group.byname... ypxfr: Exiting: Map successfully transferred Transferring services.byname... ypxfr: Exiting: Map successfully transferred Transferring rpc.bynumber... ypxfr: Exiting: Map successfully transferred Transferring rpc.byname... ypxfr: Exiting: Map successfully transferred Transferring protocols.byname... ypxfr: Exiting: Map successfully transferred Transferring master.passwd.byname... ypxfr: Exiting: Map successfully transferred Transferring networks.byname... ypxfr: Exiting: Map successfully transferred Transferring networks.byaddr... ypxfr: Exiting: Map successfully transferred Transferring netid.byname... ypxfr: Exiting: Map successfully transferred Transferring hosts.byaddr... ypxfr: Exiting: Map successfully transferred Transferring protocols.bynumber... ypxfr: Exiting: Map successfully transferred Transferring ypservers... ypxfr: Exiting: Map successfully transferred Transferring hosts.byname... ypxfr: Exiting: Map successfully transferred coltrane has been setup as an YP slave server without any errors. Remember to update map ypservers on ellington. This will generate a directory on the slave server called /var/yp/test-domain which contains copies of the NIS master server's maps. Adding these /etc/crontab entries on each slave server will force the slaves to sync their maps with the maps on the master server: 20 * * * * root /usr/libexec/ypxfr passwd.byname 21 * * * * root /usr/libexec/ypxfr passwd.byuid These entries are not mandatory because the master server automatically attempts to push any map changes to its slaves. However, since clients may depend upon the slave server to provide correct password information, it is recommended to force frequent password map updates. This is especially important on busy networks where map updates might not always complete. To finish the configuration, run /etc/netstart on the slave server in order to start the NIS services. Setting Up an <acronym>NIS</acronym> Client An NIS client binds to an NIS server using &man.ypbind.8;. This daemon broadcasts RPC requests on the local network. These requests specify the domain name configured on the client. If an NIS server in the same domain receives one of the broadcasts, it will respond to ypbind, which will record the server's address. If there are several servers available, the client will use the address of the first server to respond and will direct all of its NIS requests to that server. The client will automatically ping the server on a regular basis to make sure it is still available. If it fails to receive a reply within a reasonable amount of time, ypbind will mark the domain as unbound and begin broadcasting again in the hopes of locating another server. NIS client configuration To configure a &os; machine to be an NIS client: Edit /etc/rc.conf and add the following lines in order to set the NIS domain name and start &man.ypbind.8; during network startup: nisdomainname="test-domain" nis_client_enable="YES" To import all possible password entries from the NIS server, use vipw to remove all user accounts except one from /etc/master.passwd. When removing the accounts, keep in mind that at least one local account should remain and this account should be a member of wheel. If there is a problem with NIS, this local account can be used to log in remotely, become the superuser, and fix the problem. Before saving the edits, add the following line to the end of the file: +::::::::: This line configures the client to provide anyone with a valid account in the NIS server's password maps an account on the client. There are many ways to configure the NIS client by modifying this line. One method is described in . For more detailed reading, refer to the book Managing NFS and NIS, published by O'Reilly Media. To import all possible group entries from the NIS server, add this line to /etc/group: +:*:: To start the NIS client immediately, execute the following commands as the superuser: &prompt.root; /etc/netstart &prompt.root; service ypbind start After completing these steps, running ypcat passwd on the client should show the server's passwd map. <acronym>NIS</acronym> Security Since RPC is a broadcast-based service, any system running ypbind within the same domain can retrieve the contents of the NIS maps. To prevent unauthorized transactions, &man.ypserv.8; supports a feature called securenets which can be used to restrict access to a given set of hosts. By default, this information is stored in /var/yp/securenets, unless &man.ypserv.8; is started with and an alternate path. This file contains entries that consist of a network specification and a network mask separated by white space. Lines starting with # are considered to be comments. A sample securenets might look like this: # allow connections from local host -- mandatory 127.0.0.1 255.255.255.255 # allow connections from any host # on the 192.168.128.0 network 192.168.128.0 255.255.255.0 # allow connections from any host # between 10.0.0.0 to 10.0.15.255 # this includes the machines in the testlab 10.0.0.0 255.255.240.0 If &man.ypserv.8; receives a request from an address that matches one of these rules, it will process the request normally. If the address fails to match a rule, the request will be ignored and a warning message will be logged. If the securenets does not exist, ypserv will allow connections from any host. is an alternate mechanism for providing access control instead of securenets. While either access control mechanism adds some security, they are both vulnerable to IP spoofing attacks. All NIS-related traffic should be blocked at the firewall. Servers using securenets may fail to serve legitimate NIS clients with archaic TCP/IP implementations. Some of these implementations set all host bits to zero when doing broadcasts or fail to observe the subnet mask when calculating the broadcast address. While some of these problems can be fixed by changing the client configuration, other problems may force the retirement of these client systems or the abandonment of securenets. TCP Wrapper The use of TCP Wrapper increases the latency of the NIS server. The additional delay may be long enough to cause timeouts in client programs, especially in busy networks with slow NIS servers. If one or more clients suffer from latency, convert those clients into NIS slave servers and force them to bind to themselves. Barring Some Users In this example, the basie system is a faculty workstation within the NIS domain. The passwd map on the master NIS server contains accounts for both faculty and students. This section demonstrates how to allow faculty logins on this system while refusing student logins. To prevent specified users from logging on to a system, even if they are present in the NIS database, use vipw to add -username with the correct number of colons towards the end of /etc/master.passwd on the client, where username is the username of a user to bar from logging in. The line with the blocked user must be before the + line that allows NIS users. In this example, bill is barred from logging on to basie: basie&prompt.root; cat /etc/master.passwd root:[password]:0:0::0:0:The super-user:/root:/bin/csh toor:[password]:0:0::0:0:The other super-user:/root:/bin/sh daemon:*:1:1::0:0:Owner of many system processes:/root:/sbin/nologin operator:*:2:5::0:0:System &:/:/sbin/nologin bin:*:3:7::0:0:Binaries Commands and Source,,,:/:/sbin/nologin tty:*:4:65533::0:0:Tty Sandbox:/:/sbin/nologin kmem:*:5:65533::0:0:KMem Sandbox:/:/sbin/nologin games:*:7:13::0:0:Games pseudo-user:/usr/games:/sbin/nologin news:*:8:8::0:0:News Subsystem:/:/sbin/nologin man:*:9:9::0:0:Mister Man Pages:/usr/share/man:/sbin/nologin bind:*:53:53::0:0:Bind Sandbox:/:/sbin/nologin uucp:*:66:66::0:0:UUCP pseudo-user:/var/spool/uucppublic:/usr/libexec/uucp/uucico xten:*:67:67::0:0:X-10 daemon:/usr/local/xten:/sbin/nologin pop:*:68:6::0:0:Post Office Owner:/nonexistent:/sbin/nologin nobody:*:65534:65534::0:0:Unprivileged user:/nonexistent:/sbin/nologin -bill::::::::: +::::::::: basie&prompt.root; Using Netgroups netgroups Barring specified users from logging on to individual systems becomes unscaleable on larger networks and quickly loses the main benefit of NIS: centralized administration. Netgroups were developed to handle large, complex networks with hundreds of users and machines. Their use is comparable to &unix; groups, where the main difference is the lack of a numeric ID and the ability to define a netgroup by including both user accounts and other netgroups. To expand on the example used in this chapter, the NIS domain will be extended to add the users and systems shown in Tables 28.2 and 28.3: Additional Users User Name(s) Description alpha, beta IT department employees charlie, delta IT department apprentices echo, foxtrott, golf, ... employees able, baker, ... interns
Additional Systems Machine Name(s) Description war, death, famine, pollution Only IT employees are allowed to log onto these servers. pride, greed, envy, wrath, lust, sloth All members of the IT department are allowed to login onto these servers. one, two, three, four, ... Ordinary workstations used by employees. trashcan A very old machine without any critical data. Even interns are allowed to use this system.
When using netgroups to configure this scenario, each user is assigned to one or more netgroups and logins are then allowed or forbidden for all members of the netgroup. When adding a new machine, login restrictions must be defined for all netgroups. When a new user is added, the account must be added to one or more netgroups. If the NIS setup is planned carefully, only one central configuration file needs modification to grant or deny access to machines. The first step is the initialization of the NIS netgroup map. In &os;, this map is not created by default. On the NIS master server, use an editor to create a map named /var/yp/netgroup. This example creates four netgroups to represent IT employees, IT apprentices, employees, and interns: IT_EMP (,alpha,test-domain) (,beta,test-domain) IT_APP (,charlie,test-domain) (,delta,test-domain) USERS (,echo,test-domain) (,foxtrott,test-domain) \ (,golf,test-domain) INTERNS (,able,test-domain) (,baker,test-domain) Each entry configures a netgroup. The first column in an entry is the name of the netgroup. Each set of brackets represents either a group of one or more users or the name of another netgroup. When specifying a user, the three comma-delimited fields inside each group represent: The name of the host(s) where the other fields representing the user are valid. If a hostname is not specified, the entry is valid on all hosts. The name of the account that belongs to this netgroup. The NIS domain for the account. Accounts may be imported from other NIS domains into a netgroup. If a group contains multiple users, separate each user with whitespace. Additionally, each field may contain wildcards. See &man.netgroup.5; for details. netgroups Netgroup names longer than 8 characters should not be used. The names are case sensitive and using capital letters for netgroup names is an easy way to distinguish between user, machine and netgroup names. Some non-&os; NIS clients cannot handle netgroups containing more than 15 entries. This limit may be circumvented by creating several sub-netgroups with 15 users or fewer and a real netgroup consisting of the sub-netgroups, as seen in this example: BIGGRP1 (,joe1,domain) (,joe2,domain) (,joe3,domain) [...] BIGGRP2 (,joe16,domain) (,joe17,domain) [...] BIGGRP3 (,joe31,domain) (,joe32,domain) BIGGROUP BIGGRP1 BIGGRP2 BIGGRP3 Repeat this process if more than 225 (15 times 15) users exist within a single netgroup. To activate and distribute the new NIS map: ellington&prompt.root; cd /var/yp ellington&prompt.root; make This will generate the three NIS maps netgroup, netgroup.byhost and netgroup.byuser. Use the map key option of &man.ypcat.1; to check if the new NIS maps are available: ellington&prompt.user; ypcat -k netgroup ellington&prompt.user; ypcat -k netgroup.byhost ellington&prompt.user; ypcat -k netgroup.byuser The output of the first command should resemble the contents of /var/yp/netgroup. The second command only produces output if host-specific netgroups were created. The third command is used to get the list of netgroups for a user. To configure a client, use &man.vipw.8; to specify the name of the netgroup. For example, on the server named war, replace this line: +::::::::: with +@IT_EMP::::::::: This specifies that only the users defined in the netgroup IT_EMP will be imported into this system's password database and only those users are allowed to login to this system. This configuration also applies to the ~ function of the shell and all routines which convert between user names and numerical user IDs. In other words, cd ~user will not work, ls -l will show the numerical ID instead of the username, and find . -user joe -print will fail with the message No such user. To fix this, import all user entries without allowing them to login into the servers. This can be achieved by adding an extra line: +:::::::::/sbin/nologin This line configures the client to import all entries but to replace the shell in those entries with /sbin/nologin. Make sure that extra line is placed after +@IT_EMP:::::::::. Otherwise, all user accounts imported from NIS will have /sbin/nologin as their login shell and no one will be able to login to the system. To configure the less important servers, replace the old +::::::::: on the servers with these lines: +@IT_EMP::::::::: +@IT_APP::::::::: +:::::::::/sbin/nologin The corresponding lines for the workstations would be: +@IT_EMP::::::::: +@USERS::::::::: +:::::::::/sbin/nologin NIS supports the creation of netgroups from other netgroups which can be useful if the policy regarding user access changes. One possibility is the creation of role-based netgroups. For example, one might create a netgroup called BIGSRV to define the login restrictions for the important servers, another netgroup called SMALLSRV for the less important servers, and a third netgroup called USERBOX for the workstations. Each of these netgroups contains the netgroups that are allowed to login onto these machines. The new entries for the NIS netgroup map would look like this: BIGSRV IT_EMP IT_APP SMALLSRV IT_EMP IT_APP ITINTERN USERBOX IT_EMP ITINTERN USERS This method of defining login restrictions works reasonably well when it is possible to define groups of machines with identical restrictions. Unfortunately, this is the exception and not the rule. Most of the time, the ability to define login restrictions on a per-machine basis is required. Machine-specific netgroup definitions are another possibility to deal with the policy changes. In this scenario, the /etc/master.passwd of each system contains two lines starting with +. The first line adds a netgroup with the accounts allowed to login onto this machine and the second line adds all other accounts with /sbin/nologin as shell. It is recommended to use the ALL-CAPS version of the hostname as the name of the netgroup: +@BOXNAME::::::::: +:::::::::/sbin/nologin Once this task is completed on all the machines, there is no longer a need to modify the local versions of /etc/master.passwd ever again. All further changes can be handled by modifying the NIS map. Here is an example of a possible netgroup map for this scenario: # Define groups of users first IT_EMP (,alpha,test-domain) (,beta,test-domain) IT_APP (,charlie,test-domain) (,delta,test-domain) DEPT1 (,echo,test-domain) (,foxtrott,test-domain) DEPT2 (,golf,test-domain) (,hotel,test-domain) DEPT3 (,india,test-domain) (,juliet,test-domain) ITINTERN (,kilo,test-domain) (,lima,test-domain) D_INTERNS (,able,test-domain) (,baker,test-domain) # # Now, define some groups based on roles USERS DEPT1 DEPT2 DEPT3 BIGSRV IT_EMP IT_APP SMALLSRV IT_EMP IT_APP ITINTERN USERBOX IT_EMP ITINTERN USERS # # And a groups for a special tasks # Allow echo and golf to access our anti-virus-machine SECURITY IT_EMP (,echo,test-domain) (,golf,test-domain) # # machine-based netgroups # Our main servers WAR BIGSRV FAMINE BIGSRV # User india needs access to this server POLLUTION BIGSRV (,india,test-domain) # # This one is really important and needs more access restrictions DEATH IT_EMP # # The anti-virus-machine mentioned above ONE SECURITY # # Restrict a machine to a single user TWO (,hotel,test-domain) # [...more groups to follow] It may not always be advisable to use machine-based netgroups. When deploying a couple of dozen or hundreds of systems, role-based netgroups instead of machine-based netgroups may be used to keep the size of the NIS map within reasonable limits. Password Formats NIS password formats NIS requires that all hosts within an NIS domain use the same format for encrypting passwords. If users have trouble authenticating on an NIS client, it may be due to a differing password format. In a heterogeneous network, the format must be supported by all operating systems, where DES is the lowest common standard. To check which format a server or client is using, look at this section of /etc/login.conf: default:\ :passwd_format=des:\ :copyright=/etc/COPYRIGHT:\ [Further entries elided] In this example, the system is using the DES format. Other possible values are blf for Blowfish and md5 for MD5 encrypted passwords. If the format on a host needs to be edited to match the one being used in the NIS domain, the login capability database must be rebuilt after saving the change: &prompt.root; cap_mkdb /etc/login.conf The format of passwords for existing user accounts will not be updated until each user changes their password after the login capability database is rebuilt. Lightweight Directory Access Protocol (<acronym>LDAP</acronym>) Tom Rhodes Written by LDAP The Lightweight Directory Access Protocol (LDAP) is an application layer protocol used to access, modify, and authenticate objects using a distributed directory information service. Think of it as a phone or record book which stores several levels of hierarchical, homogeneous information. It is used in Active Directory and OpenLDAP networks and allows users to access to several levels of internal information utilizing a single account. For example, email authentication, pulling employee contact information, and internal website authentication might all make use of a single user account in the LDAP server's record base. This section provides a quick start guide for configuring an LDAP server on a &os; system. It assumes that the administrator already has a design plan which includes the type of information to store, what that information will be used for, which users should have access to that information, and how to secure this information from unauthorized access. <acronym>LDAP</acronym> Terminology and Structure LDAP uses several terms which should be understood before starting the configuration. All directory entries consist of a group of attributes. Each of these attribute sets contains a unique identifier known as a Distinguished Name (DN) which is normally built from several other attributes such as the common or Relative Distinguished Name (RDN). Similar to how directories have absolute and relative paths, consider a DN as an absolute path and the RDN as the relative path. An example LDAP entry looks like the following. This example searches for the entry for the specified user account (uid), organizational unit (ou), and organization (o): &prompt.user; ldapsearch -xb "uid=trhodes,ou=users,o=example.com" # extended LDIF # # LDAPv3 # base <uid=trhodes,ou=users,o=example.com> with scope subtree # filter: (objectclass=*) # requesting: ALL # # trhodes, users, example.com dn: uid=trhodes,ou=users,o=example.com mail: trhodes@example.com cn: Tom Rhodes uid: trhodes telephoneNumber: (123) 456-7890 # search result search: 2 result: 0 Success # numResponses: 2 # numEntries: 1 This example entry shows the values for the dn, mail, cn, uid, and telephoneNumber attributes. The cn attribute is the RDN. More information about LDAP and its terminology can be found at http://www.openldap.org/doc/admin24/intro.html. Configuring an <acronym>LDAP</acronym> Server LDAP Server &os; does not provide a built-in LDAP server. Begin the configuration by installing the net/openldap24-server package or port. Since the port has many configurable options, it is recommended that the default options are reviewed to see if the package is sufficient, and to instead compile the port if any options should be changed. In most cases, the defaults are fine. However, if SQL support is needed, this option must be enabled and the port compiled using the instructions in . Next, create the directories to hold the data and to store the certificates: &prompt.root; mkdir /var/db/openldap-data &prompt.root; mkdir /usr/local/etc/openldap/private Copy over the database configuration file: &prompt.root; cp /usr/local/etc/openldap/DB_CONFIG.example /var/db/openldap-data/DB_CONFIG The next phase is to configure the certificate authority. The following commands must be executed from /usr/local/etc/openldap/private. This is important as the file permissions need to be restrictive and users should not have access to these files. To create the certificate authority, start with this command and follow the prompts: &prompt.root; openssl req -days 365 -nodes -new -x509 -keyout ca.key -out ../ca.crt The entries for the prompts may be generic except for the Common Name. This entry must be different than the system hostname. If this will be a self signed certificate, prefix the hostname with CA for certificate authority. The next task is to create a certificate signing request and a private key. Input this command and follow the prompts: &prompt.root; openssl req -days 365 -nodes -new -keyout server.key -out server.csr During the certificate generation process, be sure to correctly set the Common Name attribute. Once complete, sign the key: &prompt.root; openssl x509 -req -days 365 -in server.csr -out ../server.crt -CA ../ca.crt -CAkey ca.key -CAcreateserial The final part of the certificate generation process is to generate and sign the client certificates: &prompt.root; openssl req -days 365 -nodes -new -keyout client.key -out client.csr &prompt.root; openssl x509 -req -days 3650 -in client.csr -out ../client.crt -CA ../ca.crt -CAkey ca.key Remember to use the same Common Name attribute when prompted. When finished, ensure that a total of eight (8) new files have been generated through the proceeding commands. If so, the next step is to edit /usr/local/etc/openldap/slapd.conf and add the following options: TLSCipherSuite HIGH:MEDIUM:+SSLv3 TLSCertificateFile /usr/local/etc/openldap/server.crt TLSCertificateKeyFile /usr/local/etc/openldap/private/server.key TLSCACertificateFile /usr/local/etc/openldap/ca.crt Then, edit /usr/local/etc/openldap/ldap.conf and add the following lines: TLS_CACERT /usr/local/etc/openldap/ca.crt TLS_CIPHER_SUITE HIGH:MEDIUM:+SSLv3 While editing this file, uncomment the following entries and set them to the desired values: , , and . Set the to contain and . Then, add two entries pointing to the certificate authority. When finished, the entries should look similar to the following: BASE dc=example,dc=com URI ldap:// ldaps:// SIZELIMIT 12 TIMELIMIT 15 TLS_CACERT /usr/local/etc/openldap/ca.crt TLS_CIPHER_SUITE HIGH:MEDIUM:+SSLv3 The default password for the server should then be changed: &prompt.root; slappasswd -h "{SHA}" >> /usr/local/etc/openldap/slapd.conf This command will prompt for the password and, if the process does not fail, a password hash will be added to the end of slapd.conf. Several hashing formats are supported. Refer to the manual page for slappasswd for more information. Next, edit /usr/local/etc/openldap/slapd.conf and add the following lines: password-hash {sha} allow bind_v2 The in this file must be updated to match the used in /usr/local/etc/openldap/ldap.conf and should also be set. A recommended value for is something like . Before saving this file, place the in front of the password output from slappasswd and delete the old . The end result should look similar to this: TLSCipherSuite HIGH:MEDIUM:+SSLv3 TLSCertificateFile /usr/local/etc/openldap/server.crt TLSCertificateKeyFile /usr/local/etc/openldap/private/server.key TLSCACertificateFile /usr/local/etc/openldap/ca.crt rootpw {SHA}W6ph5Mm5Pz8GgiULbPgzG37mj9g= Finally, enable the OpenLDAP service in /etc/rc.conf and set the URI: slapd_enable="YES" slapd_flags="-4 -h ldaps:///" At this point the server can be started and tested: &prompt.root; service slapd start If everything is configured correctly, a search of the directory should show a successful connection with a single response as in this example: &prompt.root; ldapsearch -Z # extended LDIF # # LDAPv3 # base <dc=example,dc=com> (default) with scope subtree # filter: (objectclass=*) # requesting: ALL # # search result search: 3 result: 32 No such object # numResponses: 1 If the command fails and the configuration looks correct, stop the slapd service and restart it with debugging options: &prompt.root; service slapd stop &prompt.root; /usr/local/libexec/slapd -d -1 Once the service is responding, the directory can be populated using ldapadd. In this example, a file containing this list of users is first created. Each user should use the following format: dn: dc=example,dc=com objectclass: dcObject objectclass: organization o: Example dc: Example dn: cn=Manager,dc=example,dc=com objectclass: organizationalRole cn: Manager To import this file, specify the file name. The following command will prompt for the password specified earlier and the output should look something like this: &prompt.root; ldapadd -Z -D "cn=Manager,dc=example,dc=com" -W -f import.ldif Enter LDAP Password: adding new entry "dc=example,dc=com" adding new entry "cn=Manager,dc=example,dc=com" Verify the data was added by issuing a search on the server using ldapsearch: &prompt.user; ldapsearch -Z # extended LDIF # # LDAPv3 # base <dc=example,dc=com> (default) with scope subtree # filter: (objectclass=*) # requesting: ALL # # example.com dn: dc=example,dc=com objectClass: dcObject objectClass: organization o: Example dc: Example # Manager, example.com dn: cn=Manager,dc=example,dc=com objectClass: organizationalRole cn: Manager # search result search: 3 result: 0 Success # numResponses: 3 # numEntries: 2 At this point, the server should be configured and functioning properly. Dynamic Host Configuration Protocol (<acronym>DHCP</acronym>) Dynamic Host Configuration Protocol DHCP Internet Systems Consortium (ISC) The Dynamic Host Configuration Protocol (DHCP) allows a system to connect to a network in order to be assigned the necessary addressing information for communication on that network. &os; includes the OpenBSD version of dhclient which is used by the client to obtain the addressing information. &os; does not install a DHCP server, but several servers are available in the &os; Ports Collection. The DHCP protocol is fully described in RFC 2131. Informational resources are also available at isc.org/downloads/dhcp/. This section describes how to use the built-in DHCP client. It then describes how to install and configure a DHCP server. In &os;, the &man.bpf.4; device is needed by both the DHCP server and DHCP client. This device is included in the GENERIC kernel that is installed with &os;. Users who prefer to create a custom kernel need to keep this device if DHCP is used. It should be noted that bpf also allows privileged users to run network packet sniffers on that system. Configuring a <acronym>DHCP</acronym> Client DHCP client support is included in the &os; installer, making it easy to configure a newly installed system to automatically receive its networking addressing information from an existing DHCP server. Refer to for examples of network configuration. UDP When dhclient is executed on the client machine, it begins broadcasting requests for configuration information. By default, these requests use UDP port 68. The server replies on UDP port 67, giving the client an IP address and other relevant network information such as a subnet mask, default gateway, and DNS server addresses. This information is in the form of a DHCP lease and is valid for a configurable time. This allows stale IP addresses for clients no longer connected to the network to automatically be reused. DHCP clients can obtain a great deal of information from the server. An exhaustive list may be found in &man.dhcp-options.5;. By default, when a &os; system boots, its DHCP client runs in the background, or asynchronously. Other startup scripts continue to run while the DHCP process completes, which speeds up system startup. Background DHCP works well when the DHCP server responds quickly to the client's requests. However, DHCP may take a long time to complete on some systems. If network services attempt to run before DHCP has assigned the network addressing information, they will fail. Using DHCP in synchronous mode prevents this problem as it pauses startup until the DHCP configuration has completed. This line in /etc/rc.conf is used to configure background or asynchronous mode: ifconfig_fxp0="DHCP" This line may already exist if the system was configured to use DHCP during installation. Replace the fxp0 shown in these examples with the name of the interface to be dynamically configured, as described in . To instead configure the system to use synchronous mode, and to pause during startup while DHCP completes, use SYNCDHCP: ifconfig_fxp0="SYNCDHCP" Additional client options are available. Search for dhclient in &man.rc.conf.5; for details. DHCP configuration files The DHCP client uses the following files: /etc/dhclient.conf The configuration file used by dhclient. Typically, this file contains only comments as the defaults are suitable for most clients. This configuration file is described in &man.dhclient.conf.5;. /sbin/dhclient More information about the command itself can be found in &man.dhclient.8;. /sbin/dhclient-script The &os;-specific DHCP client configuration script. It is described in &man.dhclient-script.8;, but should not need any user modification to function properly. /var/db/dhclient.leases.interface The DHCP client keeps a database of valid leases in this file, which is written as a log and is described in &man.dhclient.leases.5;. Installing and Configuring a <acronym>DHCP</acronym> Server This section demonstrates how to configure a &os; system to act as a DHCP server using the Internet Systems Consortium (ISC) implementation of the DHCP server. This implementation and its documentation can be installed using the net/isc-dhcp43-server package or port. DHCP server DHCP installation The installation of net/isc-dhcp43-server installs a sample configuration file. Copy /usr/local/etc/dhcpd.conf.example to /usr/local/etc/dhcpd.conf and make any edits to this new file. DHCP dhcpd.conf The configuration file is comprised of declarations for subnets and hosts which define the information that is provided to DHCP clients. For example, these lines configure the following: option domain-name "example.org"; option domain-name-servers ns1.example.org; option subnet-mask 255.255.255.0; default-lease-time 600; max-lease-time 72400; ddns-update-style none; subnet 10.254.239.0 netmask 255.255.255.224 { range 10.254.239.10 10.254.239.20; option routers rtr-239-0-1.example.org, rtr-239-0-2.example.org; } host fantasia { hardware ethernet 08:00:07:26:c0:a5; fixed-address fantasia.fugue.com; } This option specifies the default search domain that will be provided to clients. Refer to &man.resolv.conf.5; for more information. This option specifies a comma separated list of DNS servers that the client should use. They can be listed by their Fully Qualified Domain Names (FQDN), as seen in the example, or by their IP addresses. The subnet mask that will be provided to clients. The default lease expiry time in seconds. A client can be configured to override this value. The maximum allowed length of time, in seconds, for a lease. Should a client request a longer lease, a lease will still be issued, but it will only be valid for max-lease-time. The default of disables dynamic DNS updates. Changing this to configures the DHCP server to update a DNS server whenever it hands out a lease so that the DNS server knows which IP addresses are associated with which computers in the network. Do not change the default setting unless the DNS server has been configured to support dynamic DNS. This line creates a pool of available IP addresses which are reserved for allocation to DHCP clients. The range of addresses must be valid for the network or subnet specified in the previous line. Declares the default gateway that is valid for the network or subnet specified before the opening { bracket. Specifies the hardware MAC address of a client so that the DHCP server can recognize the client when it makes a request. Specifies that this host should always be given the same IP address. Using the hostname is correct, since the DHCP server will resolve the hostname before returning the lease information. This configuration file supports many more options. Refer to dhcpd.conf(5), installed with the server, for details and examples. Once the configuration of dhcpd.conf is complete, enable the DHCP server in /etc/rc.conf: dhcpd_enable="YES" dhcpd_ifaces="dc0" Replace the dc0 with the interface (or interfaces, separated by whitespace) that the DHCP server should listen on for DHCP client requests. Start the server by issuing the following command: &prompt.root; service isc-dhcpd start Any future changes to the configuration of the server will require the dhcpd service to be stopped and then started using &man.service.8;. The DHCP server uses the following files. Note that the manual pages are installed with the server software. DHCP configuration files /usr/local/sbin/dhcpd More information about the dhcpd server can be found in dhcpd(8). /usr/local/etc/dhcpd.conf The server configuration file needs to contain all the information that should be provided to clients, along with information regarding the operation of the server. This configuration file is described in dhcpd.conf(5). /var/db/dhcpd.leases The DHCP server keeps a database of leases it has issued in this file, which is written as a log. Refer to dhcpd.leases(5), which gives a slightly longer description. /usr/local/sbin/dhcrelay This daemon is used in advanced environments where one DHCP server forwards a request from a client to another DHCP server on a separate network. If this functionality is required, install the net/isc-dhcp43-relay package or port. The installation includes dhcrelay(8) which provides more detail. Domain Name System (<acronym>DNS</acronym>) DNS Domain Name System (DNS) is the protocol through which domain names are mapped to IP addresses, and vice versa. DNS is coordinated across the Internet through a somewhat complex system of authoritative root, Top Level Domain (TLD), and other smaller-scale name servers, which host and cache individual domain information. It is not necessary to run a name server to perform DNS lookups on a system. BIND In &os; 10, the Berkeley Internet Name Domain (BIND) has been removed from the base system and replaced with Unbound. Unbound as configured in the &os; Base is a local caching resolver. BIND is still available from The Ports Collection as dns/bind99 or dns/bind98. In &os; 9 and lower, BIND is included in &os; Base. The &os; version provides enhanced security features, a new file system layout, and automated &man.chroot.8; configuration. BIND is maintained by the Internet Systems Consortium. resolver reverse DNS root zone The following table describes some of the terms associated with DNS: <acronym>DNS</acronym> Terminology Term Definition Forward DNS Mapping of hostnames to IP addresses. Origin Refers to the domain covered in a particular zone file. named, BIND Common names for the BIND name server package within &os;. Resolver A system process through which a machine queries a name server for zone information. Reverse DNS Mapping of IP addresses to hostnames. Root zone The beginning of the Internet zone hierarchy. All zones fall under the root zone, similar to how all files in a file system fall under the root directory. Zone An individual domain, subdomain, or portion of the DNS administered by the same authority.
zones examples Examples of zones: . is how the root zone is usually referred to in documentation. org. is a Top Level Domain (TLD) under the root zone. example.org. is a zone under the org. TLD. 1.168.192.in-addr.arpa is a zone referencing all IP addresses which fall under the 192.168.1.* IP address space. As one can see, the more specific part of a hostname appears to its left. For example, example.org. is more specific than org., as org. is more specific than the root zone. The layout of each part of a hostname is much like a file system: the /dev directory falls within the root, and so on. Reasons to Run a Name Server Name servers generally come in two forms: authoritative name servers, and caching (also known as resolving) name servers. An authoritative name server is needed when: One wants to serve DNS information to the world, replying authoritatively to queries. A domain, such as example.org, is registered and IP addresses need to be assigned to hostnames under it. An IP address block requires reverse DNS entries (IP to hostname). A backup or second name server, called a slave, will reply to queries. A caching name server is needed when: A local DNS server may cache and respond more quickly than querying an outside name server. When one queries for www.FreeBSD.org, the resolver usually queries the uplink ISP's name server, and retrieves the reply. With a local, caching DNS server, the query only has to be made once to the outside world by the caching DNS server. Additional queries will not have to go outside the local network, since the information is cached locally. <acronym>DNS</acronym> Server Configuration in &os; 10.0 and Later In &os; 10.0, BIND has been replaced with Unbound. Unbound is a validating caching resolver only. If an authoritative server is needed, many are available from the Ports Collection. Unbound is provided in the &os; base system. By default, it will provide DNS resolution to the local machine only. While the base system package can be configured to provide resolution services beyond the local machine, it is recommended that such requirements be addressed by installing Unbound from the &os; Ports Collection. To enable Unbound, add the following to /etc/rc.conf: local_unbound_enable="YES" Any existing nameservers in /etc/resolv.conf will be configured as forwarders in the new Unbound configuration. If any of the listed nameservers do not support DNSSEC, local DNS resolution will fail. Be sure to test each nameserver and remove any that fail the test. The following command will show the trust tree or a failure for a nameserver running on 192.168.1.1: &prompt.user; drill -S FreeBSD.org @192.168.1.1 Once each nameserver is confirmed to support DNSSEC, start Unbound: &prompt.root; service local_unbound onestart This will take care of updating /etc/resolv.conf so that queries for DNSSEC secured domains will now work. For example, run the following to validate the FreeBSD.org DNSSEC trust tree: &prompt.user; drill -S FreeBSD.org ;; Number of trusted keys: 1 ;; Chasing: freebsd.org. A DNSSEC Trust tree: freebsd.org. (A) |---freebsd.org. (DNSKEY keytag: 36786 alg: 8 flags: 256) |---freebsd.org. (DNSKEY keytag: 32659 alg: 8 flags: 257) |---freebsd.org. (DS keytag: 32659 digest type: 2) |---org. (DNSKEY keytag: 49587 alg: 7 flags: 256) |---org. (DNSKEY keytag: 9795 alg: 7 flags: 257) |---org. (DNSKEY keytag: 21366 alg: 7 flags: 257) |---org. (DS keytag: 21366 digest type: 1) | |---. (DNSKEY keytag: 40926 alg: 8 flags: 256) | |---. (DNSKEY keytag: 19036 alg: 8 flags: 257) |---org. (DS keytag: 21366 digest type: 2) |---. (DNSKEY keytag: 40926 alg: 8 flags: 256) |---. (DNSKEY keytag: 19036 alg: 8 flags: 257) ;; Chase successful DNS Server Configuration in &os; 9.<replaceable>X</replaceable> In &os;, the BIND daemon is called named. File Description &man.named.8; The BIND daemon. &man.rndc.8; Name server control utility. /etc/namedb Directory where BIND zone information resides. /etc/namedb/named.conf Configuration file of the daemon. Depending on how a given zone is configured on the server, the files related to that zone can be found in the master, slave, or dynamic subdirectories of the /etc/namedb directory. These files contain the DNS information that will be given out by the name server in response to queries. Starting BIND BIND starting Since BIND is installed by default, configuring it is relatively simple. The default named configuration is that of a basic resolving name server, running in a &man.chroot.8; environment, and restricted to listening on the local IPv4 loopback address (127.0.0.1). To start the server one time with this configuration, use the following command: &prompt.root; service named onestart To ensure the named daemon is started at boot each time, put the following line into the /etc/rc.conf: named_enable="YES" There are many configuration options for /etc/namedb/named.conf that are beyond the scope of this document. Other startup options for named on &os; can be found in the named_* flags in /etc/defaults/rc.conf and in &man.rc.conf.5;. The section is also a good read. Configuration Files BIND configuration files Configuration files for named currently reside in /etc/namedb directory and will need modification before use unless all that is needed is a simple resolver. This is where most of the configuration will be performed. <filename>/etc/namedb/named.conf</filename> - // $FreeBSD$ + // $FreeBSD$ // // Refer to the named.conf(5) and named(8) man pages, and the documentation // in /usr/share/doc/bind9 for more details. // // If you are going to set up an authoritative server, make sure you // understand the hairy details of how DNS works. Even with // simple mistakes, you can break connectivity for affected parties, // or cause huge amounts of useless Internet traffic. options { // All file and path names are relative to the chroot directory, // if any, and should be fully qualified. directory "/etc/namedb/working"; pid-file "/var/run/named/pid"; dump-file "/var/dump/named_dump.db"; statistics-file "/var/stats/named.stats"; // If named is being used only as a local resolver, this is a safe default. // For named to be accessible to the network, comment this option, specify // the proper IP address, or delete this option. listen-on { 127.0.0.1; }; // If you have IPv6 enabled on this system, uncomment this option for // use as a local resolver. To give access to the network, specify // an IPv6 address, or the keyword "any". // listen-on-v6 { ::1; }; // These zones are already covered by the empty zones listed below. // If you remove the related empty zones below, comment these lines out. disable-empty-zone "255.255.255.255.IN-ADDR.ARPA"; disable-empty-zone "0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.IP6.ARPA"; disable-empty-zone "1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.IP6.ARPA"; // If you have a DNS server around at your upstream provider, enter // its IP address here, and enable the line below. This will make you // benefit from its cache, thus reduce overall DNS traffic in the Internet. /* forwarders { 127.0.0.1; }; */ // If the 'forwarders' clause is not empty the default is to 'forward first' // which will fall back to sending a query from your local server if the name // servers in 'forwarders' do not have the answer. Alternatively you can // force your name server to never initiate queries of its own by enabling the // following line: // forward only; // If you wish to have forwarding configured automatically based on // the entries in /etc/resolv.conf, uncomment the following line and // set named_auto_forward=yes in /etc/rc.conf. You can also enable // named_auto_forward_only (the effect of which is described above). // include "/etc/namedb/auto_forward.conf"; Just as the comment says, to benefit from an uplink's cache, forwarders can be enabled here. Under normal circumstances, a name server will recursively query the Internet looking at certain name servers until it finds the answer it is looking for. Having this enabled will have it query the uplink's name server (or name server provided) first, taking advantage of its cache. If the uplink name server in question is a heavily trafficked, fast name server, enabling this may be worthwhile. 127.0.0.1 will not work here. Change this IP address to a name server at the uplink. /* Modern versions of BIND use a random UDP port for each outgoing query by default in order to dramatically reduce the possibility of cache poisoning. All users are strongly encouraged to utilize this feature, and to configure their firewalls to accommodate it. AS A LAST RESORT in order to get around a restrictive firewall policy you can try enabling the option below. Use of this option will significantly reduce your ability to withstand cache poisoning attacks, and should be avoided if at all possible. Replace NNNNN in the example with a number between 49160 and 65530. */ // query-source address * port NNNNN; }; // If you enable a local name server, do not forget to enter 127.0.0.1 // first in your /etc/resolv.conf so this server will be queried. // Also, make sure to enable it in /etc/rc.conf. // The traditional root hints mechanism. Use this, OR the slave zones below. zone "." { type hint; file "/etc/namedb/named.root"; }; /* Slaving the following zones from the root name servers has some significant advantages: 1. Faster local resolution for your users 2. No spurious traffic will be sent from your network to the roots 3. Greater resilience to any potential root server failure/DDoS On the other hand, this method requires more monitoring than the hints file to be sure that an unexpected failure mode has not incapacitated your server. Name servers that are serving a lot of clients will benefit more from this approach than individual hosts. Use with caution. To use this mechanism, uncomment the entries below, and comment the hint zone above. As documented at http://dns.icann.org/services/axfr/ these zones: "." (the root), ARPA, IN-ADDR.ARPA, IP6.ARPA, and ROOT-SERVERS.NET are available for AXFR from these servers on IPv4 and IPv6: xfr.lax.dns.icann.org, xfr.cjr.dns.icann.org */ /* zone "." { type slave; file "/etc/namedb/slave/root.slave"; masters { 192.5.5.241; // F.ROOT-SERVERS.NET. }; notify no; }; zone "arpa" { type slave; file "/etc/namedb/slave/arpa.slave"; masters { 192.5.5.241; // F.ROOT-SERVERS.NET. }; notify no; }; */ /* Serving the following zones locally will prevent any queries for these zones leaving your network and going to the root name servers. This has two significant advantages: 1. Faster local resolution for your users 2. No spurious traffic will be sent from your network to the roots */ // RFCs 1912 and 5735 (and BCP 32 for localhost) zone "localhost" { type master; file "/etc/namedb/master/localhost-forward.db"; }; zone "127.in-addr.arpa" { type master; file "/etc/namedb/master/localhost-reverse.db"; }; zone "255.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // RFC 1912-style zone for IPv6 localhost address zone "0.ip6.arpa" { type master; file "/etc/namedb/master/localhost-reverse.db"; }; // "This" Network (RFCs 1912 and 5735) zone "0.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // Private Use Networks (RFCs 1918 and 5735) zone "10.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "16.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "17.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "18.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "19.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "20.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "21.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "22.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "23.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "24.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "25.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "26.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "27.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "28.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "29.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "30.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "31.172.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "168.192.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // Link-local/APIPA (RFCs 3927 and 5735) zone "254.169.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IETF protocol assignments (RFCs 5735 and 5736) zone "0.0.192.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // TEST-NET-[1-3] for Documentation (RFCs 5735 and 5737) zone "2.0.192.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "100.51.198.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "113.0.203.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IPv6 Range for Documentation (RFC 3849) zone "8.b.d.0.1.0.0.2.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // Domain Names for Documentation and Testing (BCP 32) zone "test" { type master; file "/etc/namedb/master/empty.db"; }; zone "example" { type master; file "/etc/namedb/master/empty.db"; }; zone "invalid" { type master; file "/etc/namedb/master/empty.db"; }; zone "example.com" { type master; file "/etc/namedb/master/empty.db"; }; zone "example.net" { type master; file "/etc/namedb/master/empty.db"; }; zone "example.org" { type master; file "/etc/namedb/master/empty.db"; }; // Router Benchmark Testing (RFCs 2544 and 5735) zone "18.198.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "19.198.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IANA Reserved - Old Class E Space (RFC 5735) zone "240.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "241.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "242.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "243.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "244.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "245.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "246.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "247.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "248.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "249.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "250.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "251.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "252.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "253.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "254.in-addr.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IPv6 Unassigned Addresses (RFC 4291) zone "1.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "3.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "4.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "5.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "6.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "7.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "8.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "9.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "a.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "b.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "c.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "d.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "e.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "0.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "1.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "2.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "3.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "4.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "5.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "6.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "7.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "8.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "9.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "a.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "b.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "0.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "1.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "2.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "3.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "4.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "5.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "6.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "7.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IPv6 ULA (RFC 4193) zone "c.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "d.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IPv6 Link Local (RFC 4291) zone "8.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "9.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "a.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "b.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IPv6 Deprecated Site-Local Addresses (RFC 3879) zone "c.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "d.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "e.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; zone "f.e.f.ip6.arpa" { type master; file "/etc/namedb/master/empty.db"; }; // IP6.INT is Deprecated (RFC 4159) zone "ip6.int" { type master; file "/etc/namedb/master/empty.db"; }; // NB: Do not use the IP addresses below, they are faked, and only // serve demonstration/documentation purposes! // // Example slave zone config entries. It can be convenient to become // a slave at least for the zone your own domain is in. Ask // your network administrator for the IP address of the responsible // master name server. // // Do not forget to include the reverse lookup zone! // This is named after the first bytes of the IP address, in reverse // order, with ".IN-ADDR.ARPA" appended, or ".IP6.ARPA" for IPv6. // // Before starting to set up a master zone, make sure you fully // understand how DNS and BIND work. There are sometimes // non-obvious pitfalls. Setting up a slave zone is usually simpler. // // NB: Do not blindly enable the examples below. :-) Use actual names // and addresses instead. /* An example dynamic zone key "exampleorgkey" { algorithm hmac-md5; secret "sf87HJqjkqh8ac87a02lla=="; }; zone "example.org" { type master; allow-update { key "exampleorgkey"; }; file "/etc/namedb/dynamic/example.org"; }; */ /* Example of a slave reverse zone zone "1.168.192.in-addr.arpa" { type slave; file "/etc/namedb/slave/1.168.192.in-addr.arpa"; masters { 192.168.1.1; }; }; */ In named.conf, these are examples of slave entries for a forward and reverse zone. For each new zone served, a new zone entry must be added to named.conf. For example, the simplest zone entry for example.org can look like: zone "example.org" { type master; file "master/example.org"; }; The zone is a master, as indicated by the statement, holding its zone information in /etc/namedb/master/example.org indicated by the statement. zone "example.org" { type slave; file "slave/example.org"; }; In the slave case, the zone information is transferred from the master name server for the particular zone, and saved in the file specified. If and when the master server dies or is unreachable, the slave name server will have the transferred zone information and will be able to serve it. Zone Files BIND zone files An example master zone file for example.org (existing within /etc/namedb/master/example.org) is as follows: $TTL 3600 ; 1 hour default TTL example.org. IN SOA ns1.example.org. admin.example.org. ( 2006051501 ; Serial 10800 ; Refresh 3600 ; Retry 604800 ; Expire 300 ; Negative Response TTL ) ; DNS Servers IN NS ns1.example.org. IN NS ns2.example.org. ; MX Records IN MX 10 mx.example.org. IN MX 20 mail.example.org. IN A 192.168.1.1 ; Machine Names localhost IN A 127.0.0.1 ns1 IN A 192.168.1.2 ns2 IN A 192.168.1.3 mx IN A 192.168.1.4 mail IN A 192.168.1.5 ; Aliases www IN CNAME example.org. Note that every hostname ending in a . is an exact hostname, whereas everything without a trailing . is relative to the origin. For example, ns1 is translated into ns1.example.org. The format of a zone file follows: recordname IN recordtype value DNS records The most commonly used DNS records: SOA start of zone authority NS an authoritative name server A a host address CNAME the canonical name for an alias MX mail exchanger PTR a domain name pointer (used in reverse DNS) example.org. IN SOA ns1.example.org. admin.example.org. ( 2006051501 ; Serial 10800 ; Refresh after 3 hours 3600 ; Retry after 1 hour 604800 ; Expire after 1 week 300 ) ; Negative Response TTL example.org. the domain name, also the origin for this zone file. ns1.example.org. the primary/authoritative name server for this zone. admin.example.org. the responsible person for this zone, email address with @ replaced. (admin@example.org becomes admin.example.org) 2006051501 the serial number of the file. This must be incremented each time the zone file is modified. Nowadays, many admins prefer a yyyymmddrr format for the serial number. 2006051501 would mean last modified 05/15/2006, the latter 01 being the first time the zone file has been modified this day. The serial number is important as it alerts slave name servers for a zone when it is updated. IN NS ns1.example.org. This is an NS entry. Every name server that is going to reply authoritatively for the zone must have one of these entries. localhost IN A 127.0.0.1 ns1 IN A 192.168.1.2 ns2 IN A 192.168.1.3 mx IN A 192.168.1.4 mail IN A 192.168.1.5 The A record indicates machine names. As seen above, ns1.example.org would resolve to 192.168.1.2. IN A 192.168.1.1 This line assigns IP address 192.168.1.1 to the current origin, in this case example.org. www IN CNAME @ The canonical name record is usually used for giving aliases to a machine. In the example, www is aliased to the master machine whose name happens to be the same as the domain name example.org (192.168.1.1). CNAMEs can never be used together with another kind of record for the same hostname. MX record IN MX 10 mail.example.org. The MX record indicates which mail servers are responsible for handling incoming mail for the zone. mail.example.org is the hostname of a mail server, and 10 is the priority of that mail server. One can have several mail servers, with priorities of 10, 20 and so on. A mail server attempting to deliver to example.org would first try the highest priority MX (the record with the lowest priority number), then the second highest, etc, until the mail can be properly delivered. For in-addr.arpa zone files (reverse DNS), the same format is used, except with PTR entries instead of A or CNAME. $TTL 3600 1.168.192.in-addr.arpa. IN SOA ns1.example.org. admin.example.org. ( 2006051501 ; Serial 10800 ; Refresh 3600 ; Retry 604800 ; Expire 300 ) ; Negative Response TTL IN NS ns1.example.org. IN NS ns2.example.org. 1 IN PTR example.org. 2 IN PTR ns1.example.org. 3 IN PTR ns2.example.org. 4 IN PTR mx.example.org. 5 IN PTR mail.example.org. This file gives the proper IP address to hostname mappings for the above fictitious domain. It is worth noting that all names on the right side of a PTR record need to be fully qualified (i.e., end in a .). Caching Name Server BIND caching name server A caching name server is a name server whose primary role is to resolve recursive queries. It simply asks queries of its own, and remembers the answers for later use. <acronym role="Domain Name Security Extensions">DNSSEC</acronym> BIND DNS security extensions Domain Name System Security Extensions, or DNSSEC for short, is a suite of specifications to protect resolving name servers from forged DNS data, such as spoofed DNS records. By using digital signatures, a resolver can verify the integrity of the record. Note that DNSSEC only provides integrity via digitally signing the Resource Records (RRs). It provides neither confidentiality nor protection against false end-user assumptions. This means that it cannot protect against people going to example.net instead of example.com. The only thing DNSSEC does is authenticate that the data has not been compromised in transit. The security of DNS is an important step in securing the Internet in general. For more in-depth details of how DNSSEC works, the relevant RFCs are a good place to start. See the list in . The following sections will demonstrate how to enable DNSSEC for an authoritative DNS server and a recursive (or caching) DNS server running BIND 9. While all versions of BIND 9 support DNSSEC, it is necessary to have at least version 9.6.2 in order to be able to use the signed root zone when validating DNS queries. This is because earlier versions lack the required algorithms to enable validation using the root zone key. It is strongly recommended to use the latest version of BIND 9.7 or later to take advantage of automatic key updating for the root key, as well as other features to automatically keep zones signed and signatures up to date. Where configurations differ between 9.6.2 and 9.7 and later, differences will be pointed out. Recursive <acronym>DNS</acronym> Server Configuration Enabling DNSSEC validation of queries performed by a recursive DNS server requires a few changes to named.conf. Before making these changes the root zone key, or trust anchor, must be acquired. Currently the root zone key is not available in a file format BIND understands, so it has to be manually converted into the proper format. The key itself can be obtained by querying the root zone for it using dig. By running &prompt.user; dig +multi +noall +answer DNSKEY . > root.dnskey the key will end up in root.dnskey. The contents should look something like this: . 93910 IN DNSKEY 257 3 8 ( AwEAAagAIKlVZrpC6Ia7gEzahOR+9W29euxhJhVVLOyQ bSEW0O8gcCjFFVQUTf6v58fLjwBd0YI0EzrAcQqBGCzh /RStIoO8g0NfnfL2MTJRkxoXbfDaUeVPQuYEhg37NZWA JQ9VnMVDxP/VHL496M/QZxkjf5/Efucp2gaDX6RS6CXp oY68LsvPVjR0ZSwzz1apAzvN9dlzEheX7ICJBBtuA6G3 LQpzW5hOA2hzCTMjJPJ8LbqF6dsV6DoBQzgul0sGIcGO Yl7OyQdXfZ57relSQageu+ipAdTTJ25AsRTAoub8ONGc LmqrAmRLKBP1dfwhYB4N7knNnulqQxA+Uk1ihz0= ) ; key id = 19036 . 93910 IN DNSKEY 256 3 8 ( AwEAAcaGQEA+OJmOzfzVfoYN249JId7gx+OZMbxy69Hf UyuGBbRN0+HuTOpBxxBCkNOL+EJB9qJxt+0FEY6ZUVjE g58sRr4ZQ6Iu6b1xTBKgc193zUARk4mmQ/PPGxn7Cn5V EGJ/1h6dNaiXuRHwR+7oWh7DnzkIJChcTqlFrXDW3tjt ) ; key id = 34525 Do not be alarmed if the obtained keys differ from this example. They might have changed since these instructions were last updated. This output actually contains two keys. The first key in the listing, with the value 257 after the DNSKEY record type, is the one needed. This value indicates that this is a Secure Entry Point (SEP), commonly known as a Key Signing Key (KSK). The second key, with value 256, is a subordinate key, commonly called a Zone Signing Key (ZSK). More on the different key types later in . Now the key must be verified and formatted so that BIND can use it. To verify the key, generate a DS RR set. Create a file containing these RRs with &prompt.user; dnssec-dsfromkey -f root.dnskey . > root.ds These records use SHA-1 and SHA-256 respectively, and should look similar to the following example, where the longer is using SHA-256. . IN DS 19036 8 1 B256BD09DC8DD59F0E0F0D8541B8328DD986DF6E . IN DS 19036 8 2 49AAC11D7B6F6446702E54A1607371607A1A41855200FD2CE1CDDE32F24E8FB5 The SHA-256 RR can now be compared to the digest in https://data.iana.org/root-anchors/root-anchors.xml. To be absolutely sure that the key has not been tampered with the data in the XML file should be verified using a proper PGP signature. Next, the key must be formatted properly. This differs a little between BIND versions 9.6.2 and 9.7 and later. In version 9.7 support was added to automatically track changes to the key and update it as necessary. This is done using managed-keys as seen in the example below. When using the older version, the key is added using a trusted-keys statement and updates must be done manually. For BIND 9.6.2 the format should look like: trusted-keys { "." 257 3 8 "AwEAAagAIKlVZrpC6Ia7gEzahOR+9W29euxhJhVVLOyQbSEW0O8gcCjF FVQUTf6v58fLjwBd0YI0EzrAcQqBGCzh/RStIoO8g0NfnfL2MTJRkxoX bfDaUeVPQuYEhg37NZWAJQ9VnMVDxP/VHL496M/QZxkjf5/Efucp2gaD X6RS6CXpoY68LsvPVjR0ZSwzz1apAzvN9dlzEheX7ICJBBtuA6G3LQpz W5hOA2hzCTMjJPJ8LbqF6dsV6DoBQzgul0sGIcGOYl7OyQdXfZ57relS Qageu+ipAdTTJ25AsRTAoub8ONGcLmqrAmRLKBP1dfwhYB4N7knNnulq QxA+Uk1ihz0="; }; For 9.7 the format will instead be: managed-keys { "." initial-key 257 3 8 "AwEAAagAIKlVZrpC6Ia7gEzahOR+9W29euxhJhVVLOyQbSEW0O8gcCjF FVQUTf6v58fLjwBd0YI0EzrAcQqBGCzh/RStIoO8g0NfnfL2MTJRkxoX bfDaUeVPQuYEhg37NZWAJQ9VnMVDxP/VHL496M/QZxkjf5/Efucp2gaD X6RS6CXpoY68LsvPVjR0ZSwzz1apAzvN9dlzEheX7ICJBBtuA6G3LQpz W5hOA2hzCTMjJPJ8LbqF6dsV6DoBQzgul0sGIcGOYl7OyQdXfZ57relS Qageu+ipAdTTJ25AsRTAoub8ONGcLmqrAmRLKBP1dfwhYB4N7knNnulq QxA+Uk1ihz0="; }; The root key can now be added to named.conf either directly or by including a file containing the key. After these steps, configure BIND to do DNSSEC validation on queries by editing named.conf and adding the following to the options directive: dnssec-enable yes; dnssec-validation yes; To verify that it is actually working use dig to make a query for a signed zone using the resolver just configured. A successful reply will contain the AD flag to indicate the data was authenticated. Running a query such as &prompt.user; dig @resolver +dnssec se ds should return the DS RR for the .se zone. In the flags: section the AD flag should be set, as seen in: ... ;; flags: qr rd ra ad; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 1 ... The resolver is now capable of authenticating DNS queries. Authoritative <acronym>DNS</acronym> Server Configuration In order to get an authoritative name server to serve a DNSSEC signed zone a little more work is required. A zone is signed using cryptographic keys which must be generated. It is possible to use only one key for this. The preferred method however is to have a strong well-protected Key Signing Key (KSK) that is not rotated very often and a Zone Signing Key (ZSK) that is rotated more frequently. Information on recommended operational practices can be found in RFC 4641: DNSSEC Operational Practices. Practices regarding the root zone can be found in DNSSEC Practice Statement for the Root Zone KSK operator and DNSSEC Practice Statement for the Root Zone ZSK operator. The KSK is used to build a chain of authority to the data in need of validation and as such is also called a Secure Entry Point (SEP) key. A message digest of this key, called a Delegation Signer (DS) record, must be published in the parent zone to establish the trust chain. How this is accomplished depends on the parent zone owner. The ZSK is used to sign the zone, and only needs to be published there. To enable DNSSEC for the example.com zone depicted in previous examples, the first step is to use dnssec-keygen to generate the KSK and ZSK key pair. This key pair can utilize different cryptographic algorithms. It is recommended to use RSA/SHA256 for the keys and 2048 bits key length should be enough. To generate the KSK for example.com, run &prompt.user; dnssec-keygen -f KSK -a RSASHA256 -b 2048 -n ZONE example.com and to generate the ZSK, run &prompt.user; dnssec-keygen -a RSASHA256 -b 2048 -n ZONE example.com dnssec-keygen outputs two files, the public and the private keys in files named similar to Kexample.com.+005+nnnnn.key (public) and Kexample.com.+005+nnnnn.private (private). The nnnnn part of the file name is a five digit key ID. Keep track of which key ID belongs to which key. This is especially important when having more than one key in a zone. It is also possible to rename the keys. For each KSK file do: &prompt.user; mv Kexample.com.+005+nnnnn.key Kexample.com.+005+nnnnn.KSK.key &prompt.user; mv Kexample.com.+005+nnnnn.private Kexample.com.+005+nnnnn.KSK.private For the ZSK files, substitute KSK for ZSK as necessary. The files can now be included in the zone file, using the $include statement. It should look something like this: $include Kexample.com.+005+nnnnn.KSK.key ; KSK $include Kexample.com.+005+nnnnn.ZSK.key ; ZSK Finally, sign the zone and tell BIND to use the signed zone file. To sign a zone dnssec-signzone is used. The command to sign the zone example.com, located in example.com.db would look similar to &prompt.user; dnssec-signzone -o example.com -k Kexample.com.+005+nnnnn.KSK example.com.db Kexample.com.+005+nnnnn.ZSK.key The key supplied to the argument is the KSK and the other key file is the ZSK that should be used in the signing. It is possible to supply more than one KSK and ZSK, which will result in the zone being signed with all supplied keys. This can be needed to supply zone data signed using more than one algorithm. The output of dnssec-signzone is a zone file with all RRs signed. This output will end up in a file with the extension .signed, such as example.com.db.signed. The DS records will also be written to a separate file dsset-example.com. To use this signed zone just modify the zone directive in named.conf to use example.com.db.signed. By default, the signatures are only valid 30 days, meaning that the zone needs to be resigned in about 15 days to be sure that resolvers are not caching records with stale signatures. It is possible to make a script and a cron job to do this. See relevant manuals for details. Be sure to keep private keys confidential, as with all cryptographic keys. When changing a key it is best to include the new key into the zone, while still signing with the old one, and then move over to using the new key to sign. After these steps are done the old key can be removed from the zone. Failure to do this might render the DNS data unavailable for a time, until the new key has propagated through the DNS hierarchy. For more information on key rollovers and other DNSSEC operational issues, see RFC 4641: DNSSEC Operational practices. Automation Using <acronym>BIND</acronym> 9.7 or Later Beginning with BIND version 9.7 a new feature called Smart Signing was introduced. This feature aims to make the key management and signing process simpler by automating parts of the task. By putting the keys into a directory called a key repository, and using the new option auto-dnssec, it is possible to create a dynamic zone which will be resigned as needed. To update this zone use nsupdate with the new option . rndc has also grown the ability to sign zones with keys in the key repository, using the option . To tell BIND to use this automatic signing and zone updating for example.com, add the following to named.conf: zone example.com { type master; key-directory "/etc/named/keys"; update-policy local; auto-dnssec maintain; file "/etc/named/dynamic/example.com.zone"; }; After making these changes, generate keys for the zone as explained in , put those keys in the key repository given as the argument to the key-directory in the zone configuration and the zone will be signed automatically. Updates to a zone configured this way must be done using nsupdate, which will take care of re-signing the zone with the new data added. For further details, see and the BIND documentation. Security Although BIND is the most common implementation of DNS, there is always the issue of security. Possible and exploitable security holes are sometimes found. While &os; automatically drops named into a &man.chroot.8; environment; there are several other security mechanisms in place which could help to lure off possible DNS service attacks. It is always good idea to read CERT's security advisories and to subscribe to the &a.security-notifications; to stay up to date with the current Internet and &os; security issues. If a problem arises, keeping sources up to date and having a fresh build of named may help. Further Reading BIND/named manual pages: &man.rndc.8; &man.named.8; &man.named.conf.5; &man.nsupdate.1; &man.dnssec-signzone.8; &man.dnssec-keygen.8; Official ISC BIND Page Official ISC BIND Forum O'Reilly DNS and BIND 5th Edition Root DNSSEC DNSSEC Trust Anchor Publication for the Root Zone RFC1034 - Domain Names - Concepts and Facilities RFC1035 - Domain Names - Implementation and Specification RFC4033 - DNS Security Introduction and Requirements RFC4034 - Resource Records for the DNS Security Extensions RFC4035 - Protocol Modifications for the DNS Security Extensions RFC4641 - DNSSEC Operational Practices RFC 5011 - Automated Updates of DNS Security (DNSSEC Trust Anchors Apache HTTP Server Murray Stokely Contributed by web servers setting up Apache The open source Apache HTTP Server is the most widely used web server. &os; does not install this web server by default, but it can be installed from the www/apache24 package or port. This section summarizes how to configure and start version 2.x of the Apache HTTP Server on &os;. For more detailed information about Apache 2.X and its configuration directives, refer to httpd.apache.org. Configuring and Starting Apache Apache configuration file In &os;, the main Apache HTTP Server configuration file is installed as /usr/local/etc/apache2x/httpd.conf, where x represents the version number. This ASCII text file begins comment lines with a #. The most frequently modified directives are: ServerRoot "/usr/local" Specifies the default directory hierarchy for the Apache installation. Binaries are stored in the bin and sbin subdirectories of the server root and configuration files are stored in the etc/apache2x subdirectory. ServerAdmin you@example.com Change this to the email address to receive problems with the server. This address also appears on some server-generated pages, such as error documents. ServerName www.example.com:80 Allows an administrator to set a hostname which is sent back to clients for the server. For example, www can be used instead of the actual hostname. If the system does not have a registered DNS name, enter its IP address instead. If the server will listen on an alternate report, change 80 to the alternate port number. DocumentRoot "/usr/local/www/apache2x/data" The directory where documents will be served from. By default, all requests are taken from this directory, but symbolic links and aliases may be used to point to other locations. It is always a good idea to make a backup copy of the default Apache configuration file before making changes. When the configuration of Apache is complete, save the file and verify the configuration using apachectl. Running apachectl configtest should return Syntax OK. Apache starting or stopping To launch Apache at system startup, add the following line to /etc/rc.conf: apache24_enable="YES" If Apache should be started with non-default options, the following line may be added to /etc/rc.conf to specify the needed flags: apache24_flags="" If apachectl does not report configuration errors, start httpd now: &prompt.root; service apache24 start The httpd service can be tested by entering http://localhost in a web browser, replacing localhost with the fully-qualified domain name of the machine running httpd. The default web page that is displayed is /usr/local/www/apache24/data/index.html. The Apache configuration can be tested for errors after making subsequent configuration changes while httpd is running using the following command: &prompt.root; service apache24 configtest It is important to note that configtest is not an &man.rc.8; standard, and should not be expected to work for all startup scripts. Virtual Hosting Virtual hosting allows multiple websites to run on one Apache server. The virtual hosts can be IP-based or name-based. IP-based virtual hosting uses a different IP address for each website. Name-based virtual hosting uses the clients HTTP/1.1 headers to figure out the hostname, which allows the websites to share the same IP address. To setup Apache to use name-based virtual hosting, add a VirtualHost block for each website. For example, for the webserver named www.domain.tld with a virtual domain of www.someotherdomain.tld, add the following entries to httpd.conf: <VirtualHost *> ServerName www.domain.tld DocumentRoot /www/domain.tld </VirtualHost> <VirtualHost *> ServerName www.someotherdomain.tld DocumentRoot /www/someotherdomain.tld </VirtualHost> For each virtual host, replace the values for ServerName and DocumentRoot with the values to be used. For more information about setting up virtual hosts, consult the official Apache documentation at: http://httpd.apache.org/docs/vhosts/. Apache Modules Apache modules Apache uses modules to augment the functionality provided by the basic server. Refer to http://httpd.apache.org/docs/current/mod/ for a complete listing of and the configuration details for the available modules. In &os;, some modules can be compiled with the www/apache24 port. Type make config within /usr/ports/www/apache24 to see which modules are available and which are enabled by default. If the module is not compiled with the port, the &os; Ports Collection provides an easy way to install many modules. This section describes three of the most commonly used modules. <filename>mod_ssl</filename> web servers secure SSL cryptography The mod_ssl module uses the OpenSSL library to provide strong cryptography via the Secure Sockets Layer (SSLv3) and Transport Layer Security (TLSv1) protocols. This module provides everything necessary to request a signed certificate from a trusted certificate signing authority to run a secure web server on &os;. In &os;, mod_ssl module is enabled by default in both the package and the port. The available configuration directives are explained at http://httpd.apache.org/docs/current/mod/mod_ssl.html. <filename>mod_perl</filename> mod_perl Perl The mod_perl module makes it possible to write Apache modules in Perl. In addition, the persistent interpreter embedded in the server avoids the overhead of starting an external interpreter and the penalty of Perl start-up time. The mod_perl can be installed using the www/mod_perl2 package or port. Documentation for using this module can be found at http://perl.apache.org/docs/2.0/index.html. <filename>mod_php</filename> Tom Rhodes Written by mod_php PHP PHP: Hypertext Preprocessor (PHP) is a general-purpose scripting language that is especially suited for web development. Capable of being embedded into HTML, its syntax draws upon C, &java;, and Perl with the intention of allowing web developers to write dynamically generated webpages quickly. To gain support for PHP5 for the Apache web server, install the www/mod_php56 package or port. This will install and configure the modules required to support dynamic PHP applications. The installation will automatically add this line to /usr/local/etc/apache24/httpd.conf: LoadModule php5_module libexec/apache24/libphp5.so Then, perform a graceful restart to load the PHP module: &prompt.root; apachectl graceful The PHP support provided by www/mod_php56 is limited. Additional support can be installed using the lang/php56-extensions port which provides a menu driven interface to the available PHP extensions. Alternatively, individual extensions can be installed using the appropriate port. For instance, to add PHP support for the MySQL database server, install databases/php56-mysql. After installing an extension, the Apache server must be reloaded to pick up the new configuration changes: &prompt.root; apachectl graceful Dynamic Websites web servers dynamic In addition to mod_perl and mod_php, other languages are available for creating dynamic web content. These include Django and Ruby on Rails. Django Python Django Django is a BSD-licensed framework designed to allow developers to write high performance, elegant web applications quickly. It provides an object-relational mapper so that data types are developed as Python objects. A rich dynamic database-access API is provided for those objects without the developer ever having to write SQL. It also provides an extensible template system so that the logic of the application is separated from the HTML presentation. Django depends on mod_python, and an SQL database engine. In &os;, the www/py-django port automatically installs mod_python and supports the PostgreSQL, MySQL, or SQLite databases, with the default being SQLite. To change the database engine, type make config within /usr/ports/www/py-django, then install the port. Once Django is installed, the application will need a project directory along with the Apache configuration in order to use the embedded Python interpreter. This interpreter is used to call the application for specific URLs on the site. To configure Apache to pass requests for certain URLs to the web application, add the following to httpd.conf, specifying the full path to the project directory: <Location "/"> SetHandler python-program PythonPath "['/dir/to/the/django/packages/'] + sys.path" PythonHandler django.core.handlers.modpython SetEnv DJANGO_SETTINGS_MODULE mysite.settings PythonAutoReload On PythonDebug On </Location> Refer to https://docs.djangoproject.com for more information on how to use Django. Ruby on Rails Ruby on Rails Ruby on Rails is another open source web framework that provides a full development stack. It is optimized to make web developers more productive and capable of writing powerful applications quickly. On &os;, it can be installed using the www/rubygem-rails package or port. Refer to http://guides.rubyonrails.org for more information on how to use Ruby on Rails. File Transfer Protocol (<acronym>FTP</acronym>) FTP servers The File Transfer Protocol (FTP) provides users with a simple way to transfer files to and from an FTP server. &os; includes FTP server software, ftpd, in the base system. &os; provides several configuration files for controlling access to the FTP server. This section summarizes these files. Refer to &man.ftpd.8; for more details about the built-in FTP server. Configuration The most important configuration step is deciding which accounts will be allowed access to the FTP server. A &os; system has a number of system accounts which should not be allowed FTP access. The list of users disallowed any FTP access can be found in /etc/ftpusers. By default, it includes system accounts. Additional users that should not be allowed access to FTP can be added. In some cases it may be desirable to restrict the access of some users without preventing them completely from using FTP. This can be accomplished be creating /etc/ftpchroot as described in &man.ftpchroot.5;. This file lists users and groups subject to FTP access restrictions. FTP anonymous To enable anonymous FTP access to the server, create a user named ftp on the &os; system. Users will then be able to log on to the FTP server with a username of ftp or anonymous. When prompted for the password, any input will be accepted, but by convention, an email address should be used as the password. The FTP server will call &man.chroot.2; when an anonymous user logs in, to restrict access to only the home directory of the ftp user. There are two text files that can be created to specify welcome messages to be displayed to FTP clients. The contents of /etc/ftpwelcome will be displayed to users before they reach the login prompt. After a successful login, the contents of /etc/ftpmotd will be displayed. Note that the path to this file is relative to the login environment, so the contents of ~ftp/etc/ftpmotd would be displayed for anonymous users. Once the FTP server has been configured, set the appropriate variable in /etc/rc.conf to start the service during boot: ftpd_enable="YES" To start the service now: &prompt.root; service ftpd start Test the connection to the FTP server by typing: &prompt.user; ftp localhost syslog log files FTP The ftpd daemon uses &man.syslog.3; to log messages. By default, the system log daemon will write messages related to FTP in /var/log/xferlog. The location of the FTP log can be modified by changing the following line in /etc/syslog.conf: ftp.info /var/log/xferlog FTP anonymous Be aware of the potential problems involved with running an anonymous FTP server. In particular, think twice about allowing anonymous users to upload files. It may turn out that the FTP site becomes a forum for the trade of unlicensed commercial software or worse. If anonymous FTP uploads are required, then verify the permissions so that these files cannot be read by other anonymous users until they have been reviewed by an administrator. File and Print Services for µsoft.windows; Clients (Samba) Samba server Microsoft Windows file server Windows clients print server Windows clients Samba is a popular open source software package that provides file and print services using the SMB/CIFS protocol. This protocol is built into µsoft.windows; systems. It can be added to non-µsoft.windows; systems by installing the Samba client libraries. The protocol allows clients to access shared data and printers. These shares can be mapped as a local disk drive and shared printers can be used as if they were local printers. On &os;, the Samba client libraries can be installed using the net/samba-smbclient port or package. The client provides the ability for a &os; system to access SMB/CIFS shares in a µsoft.windows; network. A &os; system can also be configured to act as a Samba server by installing the net/samba43 port or package. This allows the administrator to create SMB/CIFS shares on the &os; system which can be accessed by clients running µsoft.windows; or the Samba client libraries. Server Configuration Samba is configured in /usr/local/etc/smb4.conf. This file must be created before Samba can be used. A simple smb4.conf to share directories and printers with &windows; clients in a workgroup is shown here. For more complex setups involving LDAP or Active Directory, it is easier to use &man.samba-tool.8; to create the initial smb4.conf. [global] workgroup = WORKGROUP server string = Samba Server Version %v netbios name = ExampleMachine wins support = Yes security = user passdb backend = tdbsam # Example: share /usr/src accessible only to 'developer' user [src] path = /usr/src valid users = developer writable = yes browsable = yes read only = no guest ok = no public = no create mask = 0666 directory mask = 0755 Global Settings Settings that describe the network are added in /usr/local/etc/smb4.conf: workgroup The name of the workgroup to be served. netbios name The NetBIOS name by which a Samba server is known. By default, it is the same as the first component of the host's DNS name. server string The string that will be displayed in the output of net view and some other networking tools that seek to display descriptive text about the server. wins support Whether Samba will act as a WINS server. Do not enable support for WINS on more than one server on the network. Security Settings The most important settings in /usr/local/etc/smb4.conf are the security model and the backend password format. These directives control the options: security The most common settings are security = share and security = user. If the clients use usernames that are the same as their usernames on the &os; machine, user level security should be used. This is the default security policy and it requires clients to first log on before they can access shared resources. In share level security, clients do not need to log onto the server with a valid username and password before attempting to connect to a shared resource. This was the default security model for older versions of Samba. passdb backend NIS+ LDAP SQL database Samba has several different backend authentication models. Clients may be authenticated with LDAP, NIS+, an SQL database, or a modified password file. The recommended authentication method, tdbsam, is ideal for simple networks and is covered here. For larger or more complex networks, ldapsam is recommended. smbpasswd was the former default and is now obsolete. <application>Samba</application> Users &os; user accounts must be mapped to the SambaSAMAccount database for &windows; clients to access the share. Map existing &os; user accounts using &man.pdbedit.8;: &prompt.root; pdbedit -a username This section has only mentioned the most commonly used settings. Refer to the Official Samba HOWTO for additional information about the available configuration options. Starting <application>Samba</application> To enable Samba at boot time, add the following line to /etc/rc.conf: samba_enable="YES" To enable Samba4, use: samba_server_enable="YES" To start Samba now: &prompt.root; service samba start Starting SAMBA: removing stale tdbs : Starting nmbd. Starting smbd. Samba consists of three separate daemons. Both the nmbd and smbd daemons are started by samba_enable. If winbind name resolution is also required, set: winbindd_enable="YES" Samba can be stopped at any time by typing: &prompt.root; service samba stop Samba is a complex software suite with functionality that allows broad integration with µsoft.windows; networks. For more information about functionality beyond the basic configuration described here, refer to http://www.samba.org. Clock Synchronization with NTP NTP ntpd Over time, a computer's clock is prone to drift. This is problematic as many network services require the computers on a network to share the same accurate time. Accurate time is also needed to ensure that file timestamps stay consistent. The Network Time Protocol (NTP) is one way to provide clock accuracy in a network. &os; includes &man.ntpd.8; which can be configured to query other NTP servers in order to synchronize the clock on that machine or to provide time services to other computers in the network. The servers which are queried can be local to the network or provided by an ISP. In addition, an online list of publicly accessible NTP servers is available. When choosing a public NTP server, select one that is geographically close and review its usage policy. Choosing several NTP servers is recommended in case one of the servers becomes unreachable or its clock proves unreliable. As ntpd receives responses, it favors reliable servers over the less reliable ones. This section describes how to configure ntpd on &os;. Further documentation can be found in /usr/share/doc/ntp/ in HTML format. <acronym>NTP</acronym> Configuration NTP ntp.conf On &os;, the built-in ntpd can be used to synchronize a system's clock. To enable ntpd at boot time, add ntpd_enable="YES" to /etc/rc.conf. Additional variables can be specified in /etc/rc.conf. Refer to &man.rc.conf.5; and &man.ntpd.8; for details. This application reads /etc/ntp.conf to determine which NTP servers to query. Here is a simple example of an /etc/ntp.conf: Sample <filename>/etc/ntp.conf</filename> server ntplocal.example.com prefer server timeserver.example.org server ntp2a.example.net driftfile /var/db/ntp.drift The format of this file is described in &man.ntp.conf.5;. The server option specifies which servers to query, with one server listed on each line. If a server entry includes prefer, that server is preferred over other servers. A response from a preferred server will be discarded if it differs significantly from other servers' responses; otherwise it will be used. The prefer argument should only be used for NTP servers that are known to be highly accurate, such as those with special time monitoring hardware. The driftfile entry specifies which file is used to store the system clock's frequency offset. ntpd uses this to automatically compensate for the clock's natural drift, allowing it to maintain a reasonably correct setting even if it is cut off from all external time sources for a period of time. This file also stores information about previous responses from NTP servers. Since this file contains internal information for NTP, it should not be modified. By default, an NTP server is accessible to any network host. The restrict option in /etc/ntp.conf can be used to control which systems can access the server. For example, to deny all machines from accessing the NTP server, add the following line to /etc/ntp.conf: restrict default ignore This will also prevent access from other NTP servers. If there is a need to synchronize with an external NTP server, allow only that specific server. Refer to &man.ntp.conf.5; for more information. To allow machines within the network to synchronize their clocks with the server, but ensure they are not allowed to configure the server or be used as peers to synchronize against, instead use: restrict 192.168.1.0 mask 255.255.255.0 nomodify notrap where 192.168.1.0 is the local network address and 255.255.255.0 is the network's subnet mask. Multiple restrict entries are supported. For more details, refer to the Access Control Support subsection of &man.ntp.conf.5;. Once ntpd_enable="YES" has been added to /etc/rc.conf, ntpd can be started now without rebooting the system by typing: &prompt.root; service ntpd start Using <acronym>NTP</acronym> with a <acronym>PPP</acronym> Connection ntpd does not need a permanent connection to the Internet to function properly. However, if a PPP connection is configured to dial out on demand, NTP traffic should be prevented from triggering a dial out or keeping the connection alive. This can be configured with filter directives in /etc/ppp/ppp.conf. For example: set filter dial 0 deny udp src eq 123 # Prevent NTP traffic from initiating dial out set filter dial 1 permit 0 0 set filter alive 0 deny udp src eq 123 # Prevent incoming NTP traffic from keeping the connection open set filter alive 1 deny udp dst eq 123 # Prevent outgoing NTP traffic from keeping the connection open set filter alive 2 permit 0/0 0/0 For more details, refer to the PACKET FILTERING section in &man.ppp.8; and the examples in /usr/share/examples/ppp/. Some Internet access providers block low-numbered ports, preventing NTP from functioning since replies never reach the machine. <acronym>iSCSI</acronym> Initiator and Target Configuration iSCSI is a way to share storage over a network. Unlike NFS, which works at the file system level, iSCSI works at the block device level. In iSCSI terminology, the system that shares the storage is known as the target. The storage can be a physical disk, or an area representing multiple disks or a portion of a physical disk. For example, if the disk(s) are formatted with ZFS, a zvol can be created to use as the iSCSI storage. The clients which access the iSCSI storage are called initiators. To initiators, the storage available through iSCSI appears as a raw, unformatted disk known as a LUN. Device nodes for the disk appear in /dev/ and the device must be separately formatted and mounted. Beginning with 10.0-RELEASE, &os; provides a native, kernel-based iSCSI target and initiator. This section describes how to configure a &os; system as a target or an initiator. Configuring an <acronym>iSCSI</acronym> Target The native iSCSI target is supported starting with &os; 10.0-RELEASE. To use iSCSI in older versions of &os;, install a userspace target from the Ports Collection, such as net/istgt. This chapter only describes the native target. To configure an iSCSI target, create the /etc/ctl.conf configuration file, add a line to /etc/rc.conf to make sure the &man.ctld.8; daemon is automatically started at boot, and then start the daemon. The following is an example of a simple /etc/ctl.conf configuration file. Refer to &man.ctl.conf.5; for a more complete description of this file's available options. portal-group pg0 { discovery-auth-group no-authentication listen 0.0.0.0 listen [::] } target iqn.2012-06.com.example:target0 { auth-group no-authentication portal-group pg0 lun 0 { path /data/target0-0 size 4G } } The first entry defines the pg0 portal group. Portal groups define which network addresses the &man.ctld.8; daemon will listen on. The discovery-auth-group no-authentication entry indicates that any initiator is allowed to perform iSCSI target discovery without authentication. Lines three and four configure &man.ctld.8; to listen on all IPv4 (listen 0.0.0.0) and IPv6 (listen [::]) addresses on the default port of 3260. It is not necessary to define a portal group as there is a built-in portal group called default. In this case, the difference between default and pg0 is that with default, target discovery is always denied, while with pg0, it is always allowed. The second entry defines a single target. Target has two possible meanings: a machine serving iSCSI or a named group of LUNs. This example uses the latter meaning, where iqn.2012-06.com.example:target0 is the target name. This target name is suitable for testing purposes. For actual use, change com.example to the real domain name, reversed. The 2012-06 represents the year and month of acquiring control of that domain name, and target0 can be any value. Any number of targets can be defined in this configuration file. The auth-group no-authentication line allows all initiators to connect to the specified target and portal-group pg0 makes the target reachable through the pg0 portal group. The next section defines the LUN. To the initiator, each LUN will be visible as a separate disk device. Multiple LUNs can be defined for each target. Each LUN is identified by a number, where LUN 0 is mandatory. The path /data/target0-0 line defines the full path to a file or zvol backing the LUN. That path must exist before starting &man.ctld.8;. The second line is optional and specifies the size of the LUN. Next, to make sure the &man.ctld.8; daemon is started at boot, add this line to /etc/rc.conf: ctld_enable="YES" To start &man.ctld.8; now, run this command: &prompt.root; service ctld start As the &man.ctld.8; daemon is started, it reads /etc/ctl.conf. If this file is edited after the daemon starts, use this command so that the changes take effect immediately: &prompt.root; service ctld reload Authentication The previous example is inherently insecure as it uses no authentication, granting anyone full access to all targets. To require a username and password to access targets, modify the configuration as follows: auth-group ag0 { chap username1 secretsecret chap username2 anothersecret } portal-group pg0 { discovery-auth-group no-authentication listen 0.0.0.0 listen [::] } target iqn.2012-06.com.example:target0 { auth-group ag0 portal-group pg0 lun 0 { path /data/target0-0 size 4G } } The auth-group section defines username and password pairs. An initiator trying to connect to iqn.2012-06.com.example:target0 must first specify a defined username and secret. However, target discovery is still permitted without authentication. To require target discovery authentication, set discovery-auth-group to a defined auth-group name instead of no-authentication. It is common to define a single exported target for every initiator. As a shorthand for the syntax above, the username and password can be specified directly in the target entry: target iqn.2012-06.com.example:target0 { portal-group pg0 chap username1 secretsecret lun 0 { path /data/target0-0 size 4G } } Configuring an <acronym>iSCSI</acronym> Initiator The iSCSI initiator described in this section is supported starting with &os; 10.0-RELEASE. To use the iSCSI initiator available in older versions, refer to &man.iscontrol.8;. The iSCSI initiator requires that the &man.iscsid.8; daemon is running. This daemon does not use a configuration file. To start it automatically at boot, add this line to /etc/rc.conf: iscsid_enable="YES" To start &man.iscsid.8; now, run this command: &prompt.root; service iscsid start Connecting to a target can be done with or without an /etc/iscsi.conf configuration file. This section demonstrates both types of connections. Connecting to a Target Without a Configuration File To connect an initiator to a single target, specify the IP address of the portal and the name of the target: &prompt.root; iscsictl -A -p 10.10.10.10 -t iqn.2012-06.com.example:target0 To verify if the connection succeeded, run iscsictl without any arguments. The output should look similar to this: Target name Target portal State iqn.2012-06.com.example:target0 10.10.10.10 Connected: da0 In this example, the iSCSI session was successfully established, with /dev/da0 representing the attached LUN. If the iqn.2012-06.com.example:target0 target exports more than one LUN, multiple device nodes will be shown in that section of the output: Connected: da0 da1 da2. Any errors will be reported in the output, as well as the system logs. For example, this message usually means that the &man.iscsid.8; daemon is not running: Target name Target portal State iqn.2012-06.com.example:target0 10.10.10.10 Waiting for iscsid(8) The following message suggests a networking problem, such as a wrong IP address or port: Target name Target portal State iqn.2012-06.com.example:target0 10.10.10.11 Connection refused This message means that the specified target name is wrong: Target name Target portal State iqn.2012-06.com.example:target0 10.10.10.10 Not found This message means that the target requires authentication: Target name Target portal State iqn.2012-06.com.example:target0 10.10.10.10 Authentication failed To specify a CHAP username and secret, use this syntax: &prompt.root; iscsictl -A -p 10.10.10.10 -t iqn.2012-06.com.example:target0 -u user -s secretsecret Connecting to a Target with a Configuration File To connect using a configuration file, create /etc/iscsi.conf with contents like this: t0 { TargetAddress = 10.10.10.10 TargetName = iqn.2012-06.com.example:target0 AuthMethod = CHAP chapIName = user chapSecret = secretsecret } The t0 specifies a nickname for the configuration file section. It will be used by the initiator to specify which configuration to use. The other lines specify the parameters to use during connection. The TargetAddress and TargetName are mandatory, whereas the other options are optional. In this example, the CHAP username and secret are shown. To connect to the defined target, specify the nickname: &prompt.root; iscsictl -An t0 Alternately, to connect to all targets defined in the configuration file, use: &prompt.root; iscsictl -Aa To make the initiator automatically connect to all targets in /etc/iscsi.conf, add the following to /etc/rc.conf: iscsictl_enable="YES" iscsictl_flags="-Aa"