[LU-2078] sync_on_lock_cancel too aggressive

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.4.0
• Labels: server topsequoia
• Severity: 3

Epic

server

• Project: Orion
• Assignee: Alex Zhuravlev
• Reporter: Brian Behlendorf

Description

While running some single shared file IOR tests I noticed, with 'zpool iostat', that there were a lots write going on during the read phase. This was absolutely destroying performance and it wasn't at all clear to me why there should be any reads at all. IOR is just reading during this phase, no writes.

On a hunch I set 'sync_on_lock_cancel=never' and the writes stopped entirely and performance improved at least 6x. I haven't looked at the code yet but I think there might be a few issues here.

1) Why were so many lock revocations occurring during the read phase. I would have thought that each client would end up with a single concurrent read lock rather quickly. However, this behavior persistent for the 30 minutes it took to read all the data back.

2) When revoking a read lock, or even a write lock which isn't actually covering any dirty data, there's no reason to issue the sync. I suspect this was never noticed because under ldiskfs if there's no dirty data to sync performing the sync is basically free. Under ZFS however if you call sync you will force the current txg to be written out. That will result in at a minimum the vdev labels at the beginning and end of each disk to be written. Lots of syncs will result each disk head basically seeking from the being to end of the disk repeatedly. Very very bad.

Issue Links

• is related to: LU-2085 sanityn test_16 (fsx) ran over its Autotest time

Comments

[Ian Colle (Inactive)]

Patch landed.

[Christopher Morrone (Inactive)]

Patch landed on master: http://review.whamcloud.com/4117

[Prakash Surya (Inactive)]

I went ahead and ran the same test with `sync_on_lock_cancel=never` as well. Here's the logs and performance numbers.

Single Shared File Read Performance (MiB/s)

        |  always |  never  |
      --+---------+---------+
58chaos | 1311.51 | 1265.46 |
      --+---------+---------+

[Prakash Surya (Inactive)]

Same logs as before, except with our 58chaos tag (reverted debug patch-set 1 and applied debug patch-set 2 as requested).

With this tag, I saw the best performance numbers of all my tests run so far (roughly equal to the sync_on_lock_cancel=never tests):

Single Shared File Read Performance (MiB/s)

        |  always |   never  |
      --+---------+----------+
58chaos | 1311.51 | untested |
      --+---------+----------+

So my question is, can we safely drop the write lock as is done in the debug patch? Or is this just a proof of concept?

Also, What exactly is that lock protecting, and why?

[Alex Zhuravlev]

thanks.. here is another patch: http://review.whamcloud.com/#change,3456 hope this time we'll get some improvement. again, a log with +dlmtrace +rpctrace is very welcome. thanks in advance.

[Prakash Surya (Inactive)]

This has the IOR output from the read and write phase of the test. Along with Lustre Logs from the OSS with +dlmtrace and +rpctrace.

[Alex Zhuravlev]

thanks for the data Prakash. I'm still looking at dlmtrace.. it's still not clear why we don't reach
the peak performance: the very first PR lock should commit the whole object and all subsequent reads
should skip osd_sync().. so here is a debug patch: http://review.whamcloud.com/#change,3456
please try, just current orion + this patch with sync_on_cancel=always should be enough. a log with
+dlmtrace +rpctrace would be very useful. thanks in advance.

[Prakash Surya (Inactive)]

IOR output for the read and write phase of each test. And MDS and OSS server logs for the read phase of each test with '+dlmtrace' enabled.