diff --git a/sbin/route/route.8 b/sbin/route/route.8 index afcf55ab44c7..25aaf8087343 100644 --- a/sbin/route/route.8 +++ b/sbin/route/route.8 @@ -1,556 +1,568 @@ .\" Copyright (c) 1983, 1991, 1993 .\" The Regents of the University of California. All rights reserved. .\" .\" Redistribution and use in source and binary forms, with or without .\" modification, are permitted provided that the following conditions .\" are met: .\" 1. Redistributions of source code must retain the above copyright .\" notice, this list of conditions and the following disclaimer. .\" 2. Redistributions in binary form must reproduce the above copyright .\" notice, this list of conditions and the following disclaimer in the .\" documentation and/or other materials provided with the distribution. .\" 3. Neither the name of the University nor the names of its contributors .\" may be used to endorse or promote products derived from this software .\" without specific prior written permission. .\" .\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND .\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE .\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE .\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE .\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL .\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS .\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) .\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT .\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY .\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF .\" SUCH DAMAGE. .\" .\" @(#)route.8 8.3 (Berkeley) 3/19/94 .\" $FreeBSD$ .\" -.Dd March 14, 2023 +.Dd May 27, 2023 .Dt ROUTE 8 .Os .Sh NAME .Nm route .Nd manually manipulate the routing tables .Sh SYNOPSIS .Nm .Op Fl dnqtv .Ar command .Oo .Op Ar modifiers .Ar args .Oc .Sh DESCRIPTION The .Nm utility is used to manually manipulate the network routing tables. It normally is not needed, as a system routing table management daemon, such as .Xr routed 8 , should tend to this task. .Pp The .Nm utility supports a limited number of general options, but a rich command language, enabling the user to specify any arbitrary request that could be delivered via the programmatic interface discussed in .Xr route 4 . .Pp The following options are available: .Bl -tag -width indent .It Fl 4 Specify .Cm inet address family as family hint for subcommands. .It Fl 6 Specify .Cm inet6 address family as family hint for subcommands. .It Fl d Run in debug-only mode, i.e., do not actually modify the routing table. .It Fl n Bypass attempts to print host and network names symbolically when reporting actions. (The process of translating between symbolic names and numerical equivalents can be quite time consuming, and may require correct operation of the network; thus it may be expedient to forget this, especially when attempting to repair networking operations). .It Fl t Run in test-only mode. .Pa /dev/null is used instead of a socket. .It Fl v (verbose) Print additional details. .It Fl q Suppress all output from the .Cm add , change , delete , and .Cm flush commands. .El .Pp The .Nm utility provides the following commands: .Pp .Bl -tag -width Fl -compact .It Cm add Add a route. .It Cm flush Remove all routes. .It Cm delete Delete a specific route. .It Cm del Another name for the .Cm delete command. .It Cm change Change aspects of a route (such as its gateway). .It Cm get Lookup and display the route for a destination. .It Cm monitor Continuously report any changes to the routing information base, routing lookup misses, or suspected network partitionings. .It Cm show Another name for the .Cm get command. .El .Pp The monitor command has the syntax: .Pp .Bd -ragged -offset indent -compact .Nm .Op Fl n .Cm monitor Op Fl fib Ar number .Ed .Pp The flush command has the syntax: .Pp .Bd -ragged -offset indent -compact .Nm .Op Fl n .Cm flush Oo Ar family Oc Op Fl fib Ar number .Ed .Pp If the .Cm flush command is specified, .Nm will ``flush'' the routing tables of all gateway entries. When the address family may is specified by any of the .Fl inet6 , or .Fl inet modifiers, only routes having destinations with addresses in the delineated family will be deleted. Additionally, .Fl 4 or .Fl 6 can be used as aliases for .Fl inet and .Fl inet6 modifiers. When a .Fl fib option is specified, the operation will be applied to the specified FIB .Pq routing table . .Pp The add command has the following syntax: .Pp .Bd -ragged -offset indent -compact .Nm .Op Fl n .Cm add .Op Fl net No \&| Fl host .Ar destination gateway .Op Ar netmask .Op Fl fib Ar number .Ed .Pp and the other commands have the following syntax: .Pp .Bd -ragged -offset indent -compact .Nm .Op Fl n .Ar command .Op Fl net No \&| Fl host .Ar destination .Op Ar gateway Op Ar netmask .Op Fl fib Ar number .Ed .Pp where .Ar destination is the destination host or network, .Ar gateway is the next-hop intermediary via which packets should be routed. Routes to a particular host may be distinguished from those to a network by interpreting the Internet address specified as the .Ar destination argument. The optional modifiers .Fl net and .Fl host force the destination to be interpreted as a network or a host, respectively. Otherwise, if the .Ar destination has a .Dq local address part of INADDR_ANY .Pq Li 0.0.0.0 , or if the .Ar destination is the symbolic name of a network, then the route is assumed to be to a network; otherwise, it is presumed to be a route to a host. Optionally, the .Ar destination could also be specified in the .Ar net Ns / Ns Ar bits format. .Pp For example, .Li 128.32 is interpreted as .Fl host Li 128.0.0.32 ; .Li 128.32.130 is interpreted as .Fl host Li 128.32.0.130 ; .Fl net Li 128.32 is interpreted as .Li 128.32.0.0 ; .Fl net Li 128.32.130 is interpreted as .Li 128.32.130.0 ; and .Li 192.168.64/20 is interpreted as .Fl net Li 192.168.64 Fl netmask Li 255.255.240.0 . .Pp A .Ar destination of .Ar default is a synonym for the default route. For .Li IPv4 it is .Fl net Fl inet Li 0.0.0.0 , and for .Li IPv6 it is .Fl net Fl inet6 Li :: . .Pp If the destination is directly reachable via an interface requiring no intermediary system to act as a gateway, the .Fl interface modifier should be specified; the gateway given is the address of this host on the common network, indicating the interface to be used for transmission. Alternately, if the interface is point to point the name of the interface itself may be given, in which case the route remains valid even if the local or remote addresses change. .Pp The optional .Fl netmask modifier is intended to achieve the effect of an OSI ESIS redirect with the netmask option, or to manually add subnet routes with netmasks different from that of the implied network interface (as would otherwise be communicated using the OSPF or ISIS routing protocols). One specifies an additional ensuing address parameter (to be interpreted as a network mask). The implicit network mask generated in the AF_INET case can be overridden by making sure this option follows the destination parameter. .Pp For .Dv AF_INET6 , the .Fl prefixlen qualifier is available instead of the .Fl mask qualifier because non-continuous masks are not allowed in IPv6. For example, .Fl prefixlen Li 32 specifies that a network mask of .Li ffff:ffff:0000:0000:0000:0000:0000:0000 will be used. The default prefixlen is 64. However, it is assumed to be 0 if .Cm default is specified for .Ar destination . Note that the qualifier works only for .Dv AF_INET6 address family. .Pp Routes have associated flags which influence operation of the protocols when sending to destinations matched by the routes. These flags may be set (or sometimes cleared) by indicating the following corresponding modifiers: .Bd -literal -xresolve RTF_XRESOLVE - emit mesg on use (for external lookup) -iface ~RTF_GATEWAY - destination is directly reachable -static RTF_STATIC - manually added route -nostatic ~RTF_STATIC - pretend route added by kernel or daemon -reject RTF_REJECT - emit an ICMP unreachable when matched -blackhole RTF_BLACKHOLE - silently discard pkts (during updates) -proto1 RTF_PROTO1 - set protocol specific routing flag #1 -proto2 RTF_PROTO2 - set protocol specific routing flag #2 .Ed .Pp The optional modifiers .Fl rtt , .Fl rttvar , .Fl sendpipe , .Fl recvpipe , .Fl mtu , .Fl hopcount , .Fl expire , and .Fl ssthresh provide initial values to quantities maintained in the routing entry by transport level protocols, such as TCP or TP4. These may be individually locked by preceding each such modifier to be locked by the .Fl lock meta-modifier, or one can specify that all ensuing metrics may be locked by the .Fl lockrest meta-modifier. .Pp Note that .Fl expire accepts expiration time of the route as the number of seconds since the Epoch .Pq see Xr time 3 . When the first character of the number is .Dq + or .Dq - , it is interpreted as a value relative to the current time. .Pp The optional modifier .Fl fib Ar number specifies that the command will be applied to a non-default FIB. The .Ar number must be smaller than the .Va net.fibs .Xr sysctl 8 MIB. When this modifier is not specified, or a negative number is specified, the default FIB shown in the .Va net.my_fibnum .Xr sysctl 8 MIB will be used. .Pp The .Ar number allows multiple FIBs by a comma-separeted list and/or range specification. The .Qq Fl fib Li 2,4,6 means the FIB number 2, 4, and 6. The .Qq Fl fib Li 1,3-5,6 means the 1, 3, 4, 5, and 6. .Pp In a .Cm change or .Cm add command where the destination and gateway are not sufficient to specify the route (as in the ISO case where several interfaces may have the same address), the .Fl ifp or .Fl ifa modifiers may be used to determine the interface or interface address. .Pp All symbolic names specified for a .Ar destination or .Ar gateway are looked up first as a host name using .Xr gethostbyname 3 . If this lookup fails, .Xr getnetbyname 3 is then used to interpret the name as that of a network. .Pp The .Nm utility uses a routing socket and the new message types .Dv RTM_ADD , RTM_DELETE , RTM_GET , and .Dv RTM_CHANGE . As such, only the super-user may modify the routing tables. .Pp .Fx provides support for scalable multipath routing. It is activated by default, but can be turned off by setting the .Va net.route.multipath .Xr sysctl 8 MIB to 0. .Pp There are multiple route lookup algorithms available. They can be configured by setting .Va net.route.algo.inet.algo for IPv4 and .Va net.route.algo.inet6.algo for IPv6 .Xr sysctl 8 MIBs. .Pp A list of available algorithms can be obtained by accessing the following .Xr sysctl 8 MIBs .Va net.route.algo.inet.algo_list for IPv4 and .Va net.route.algo.inet6.algo_list for IPv6. .Pp The following algorithms are available: .Bl -tag -width radix_lockless .It radix Base system radix backend. .It bsearch Lockless binary search in a special IP array, tailored for a small FIB with <16 routes. This algorithm is only available for IPv4. .It radix_lockless Lockless immutable radix, re-created on every rtable change, tailored for a small FIB with <1000 routes. .It dpdk_lpm DPDK DIR24-8-based lookups, lockless datastructure, optimized -for a large FIB. +for large FIBs. +DIR24-8 relies on a large flat lookup table (64 MB with IPv4) which is +directly indexed by the more significant portion of the lookup key. In order to use the dpdk_lpm algorithm one or both of the following kernel modules must be loaded via .Xr loader.conf 5 : .Bl -tag -width dpdk_lpm6.ko -compact .It dpdk_lpm4.ko DPDK implementation for IPv4. .It dpdk_lpm6.ko DPDK implementation for IPv6. .El +.It dxr +IPv4 only, lockless, compressed lookup structure +(below 2.5 Bytes per IPv4 prefix for large BGP FIBs) +which easily fits into modern CPU cache hierarchies, +lookup throughput scales linearly with CPU cores. +Loadable as a kernel module at runtime or via +.Xr loader.conf 5 : +.Bl -tag -width fib_dxr.ko -compact +.It fib_dxr.ko +.El .El .Pp The algorithms are selected automatically based on the size of the routing table of the system. They can be changed, but not every algorithm performs best for every FIB size. .Sh EXIT STATUS .Ex -std .Sh EXAMPLES Add a default route to the network routing table. This will send all packets for destinations not available in the routing table to the default gateway at 192.168.1.1: .Pp .Dl route add -net 0.0.0.0/0 192.168.1.1 .Pp A shorter version of adding a default route can also be written as: .Pp .Dl route add default 192.168.1.1 .Pp Add a static route to the 172.16.10.0/24 network via the 172.16.1.1 gateway: .Pp .Dl route add -net 172.16.10.0/24 172.16.1.1 .Pp Change the gateway of an already established static route in the routing table: .Pp .Dl route change -net 172.16.10.0/24 172.16.1.2 .Pp Display the route for a destination network: .Pp .Dl route show 172.16.10.0 .Pp Delete a static route from the routing table: .Pp .Dl route delete -net 172.16.10.0/24 172.16.1.2 .Pp Remove all routes from the routing table: .Pp .Dl route flush .Pp The routing table can be listed with .Xr netstat 1 . .Sh DIAGNOSTICS .Bl -diag .It "add [host \&| network ] %s: gateway %s flags %x" The specified route is being added to the tables. The values printed are from the routing table entry supplied in the .Xr ioctl 2 call. If the gateway address used was not the primary address of the gateway (the first one returned by .Xr gethostbyname 3 ) , the gateway address is printed numerically as well as symbolically. .It "delete [ host \&| network ] %s: gateway %s flags %x" As above, but when deleting an entry. .It "%s %s done" When the .Cm flush command is specified, each routing table entry deleted is indicated with a message of this form. .It "Network is unreachable" An attempt to add a route failed because the gateway listed was not on a directly-connected network. The next-hop gateway must be given. .It "not in table" A delete operation was attempted for an entry which was not present in the tables. .It "routing table overflow" An add operation was attempted, but the system was low on resources and was unable to allocate memory to create the new entry. .It "gateway uses the same route" A .Cm change operation resulted in a route whose gateway uses the same route as the one being changed. The next-hop gateway should be reachable through a different route. .El .Sh SEE ALSO .Xr netstat 1 , .Xr netintro 4 , .Xr route 4 , .Xr loader.conf 5 , .Xr arp 8 , .Xr routed 8 .Sh HISTORY The .Nm utility appeared in .Bx 4.2 . .Sh BUGS The first paragraph may have slightly exaggerated .Xr routed 8 Ns 's abilities. .Pp Currently, routes with the .Dv RTF_BLACKHOLE flag set need to have the gateway set to an instance of the .Xr lo 4 driver, using the .Fl iface option, for the flag to have any effect; unless IP fast forwarding is enabled, in which case the meaning of the flag will always be honored.