Page MenuHomeFreeBSD

camProject
ActivePublic

Watchers

  • This project does not have any watchers.
  • View All

Details

Description

CAM: Common Access Method

Recent Activity

Today

mav added a comment to D26912: RFC: Disk I/O priority support.

I've tried opposite approach of adding LOWPRIO flag instead and using it only for background operations in few places, and marking BIOs without it high-priority in ATA/SCSI. But while testing it I've noticed that disk random IOPS drop to almost non-NCQ level on a mix of different priorities. And I am measuring the same on both WD and HGST. I don't understand what is going on there, may be I am missing something, but that is unacceptable trade-off to me. I've uploaded my present patch in case somebody wish to play, but probably won't commit it in this state.

Thu, Oct 29, 5:44 PM · cam
mav updated the diff for D26912: RFC: Disk I/O priority support.
Thu, Oct 29, 5:22 PM · cam

Yesterday

mav added a comment to D26912: RFC: Disk I/O priority support.

Priority is working on top of tag, affecting specifically only commands tagged as SIMPLE . The ORDERED and HEAD tags still have their function as they are mandatory in their fencing semantics, while priority is a softer hint for a schduler.

Wed, Oct 28, 8:26 PM · cam
scottl added a comment to D26912: RFC: Disk I/O priority support.

I think that the intention of the feature from the manufacturers is for background sync and scrub workloads, not filesystem consistency operations. Regarding SAS, tags have always been the mechanism for setting priority and creating barriers. Ordered tags, head-of-queue tags, etc. mpr/mps support these, but they're largely unused because BIO_ORDERED was removed from FreeBSD. I'm not aware of SAS adopting the same SATA priority scheme.

Wed, Oct 28, 8:21 PM · cam
mav added a comment to D26912: RFC: Disk I/O priority support.

More experiments with SATA WD REDs show that priorities there more like absolute with deadline. On WD20EFRX-68E on heavy random workload I see low-priority requests in presence of high-priority are all delayed for about a second, while on WD80EFZX-68U they are all delayed for about 5 seconds. So big difference makes me think it is unusable for differentiation of sync vs async requests, but should still be good for read/write vs scrub/initialization/etc differentiation. Unfortunately I still haven't found any capable SAS drive to check there, but considering SATL directly map one into another I suppose they should have the same (absolute) semantics.

Wed, Oct 28, 8:01 PM · cam

Sat, Oct 24

imp added a comment to D26912: RFC: Disk I/O priority support.

I could see a use for at least three levels of priority: low priority (default) for asynchronous I/O, mid-level priority for synchronous reads, high priority for synchronous writes.

Sat, Oct 24, 12:17 AM · cam

Fri, Oct 23

mckusick added a comment to D26912: RFC: Disk I/O priority support.

It would be trivial to request high priority for synchronous writes in bwrite() and if desired synchronous reads in bread(). That would have effects for several filesystems.

Fri, Oct 23, 6:16 AM · cam
imp added a comment to D26912: RFC: Disk I/O priority support.

Interesting. I have patches to iosched that marks metadata requests and topqueues them, but doesn't try to prioritize in the drive. It doesn't handle writes, though (we don't need them, but it's one of the reasons I've not conmitted)... it gives a modest boost to open latency, but not as much as the async open chuck is working on.

Fri, Oct 23, 4:10 AM · cam

Thu, Oct 22

cem updated subscribers of D26912: RFC: Disk I/O priority support.
Thu, Oct 22, 11:22 PM · cam
mav requested review of D26912: RFC: Disk I/O priority support.
Thu, Oct 22, 9:23 PM · cam

Sat, Oct 17

imp added inline comments to D25476: Fix use after free panic and state transitions in mps(4) and mpr(4).
Sat, Oct 17, 3:26 PM · cam
imp added a comment to D25476: Fix use after free panic and state transitions in mps(4) and mpr(4).

I've had good luck running this for 2 or 3 firmware cycles now at Netflix... We're down 3/4 in panics, and at least part of the remaining 1/4 appear to be due to a small mismerge of two files so they were out of sync with the rest of the state machine....

Sat, Oct 17, 3:08 PM · cam

Jul 21 2020

imp added a comment to D25476: Fix use after free panic and state transitions in mps(4) and mpr(4).

I'm deploying this to one or two of the machines that we see panics from this every few days.

Jul 21 2020, 5:07 PM · cam

Jun 30 2020

imp accepted D25476: Fix use after free panic and state transitions in mps(4) and mpr(4).

I only looked at mpr, but this looks good to me. It's good you have a way to recreate it. I have random machines failing with this (about ~0.001%/day), but can't find one to recreate it. I like this approach, and is similar to the one I took with other commands and target reset. thanks for fixing.

Jun 30 2020, 4:33 PM · cam

Jun 26 2020

ken requested review of D25476: Fix use after free panic and state transitions in mps(4) and mpr(4).
Jun 26 2020, 8:53 PM · cam

Apr 22 2020

lwhsu created cam.
Apr 22 2020, 5:12 AM