[LU-18728] sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.17.0
• Labels: zfs-2.2.7
• Severity: 3
• Assignee: Bruno Faccini
• Reporter: Bruno Faccini

Description

sanity/test_413a failure/timeout example at https://testing.whamcloud.com/test_sets/2398dece-98fe-44d8-b89d-3c76dbef0216

conf-sanity/test_32[b,c,e,h] failures example at https://testing.whamcloud.com/test_sets/431e1d95-39e3-43bb-b721-df2e51d0ea6f

all look directly related to LU-18646 findings ("dd if=/dev/zero ..." commands do not consume expected space due to ZFS recently using data compression by default) and could be fixed the same way (https://review.whamcloud.com/c/fs/lustre-release/+/57871).

there are more conf-sanity subtests failures following but they do not look directly caused by this same problem and may only be consequences of the first failures...

Issue Links

• is duplicated by: LU-18729 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18730 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18731 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18732 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18733 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18734 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is duplicated by: LU-18735 sanity/test_413a and conf-sanity/test_32[b,c,e,h] permanent timeout/failures with ZFS back-end only
• is related to: LU-18387 RHEL 9.5 support
• is related to: LU-18652 conf-sanity test_32b: FAIL: verify quota failed
• is related to: LU-18646 ost-pools test_23b: FAIL: dd didn't fail with ENOSPC (26214400 > 22639616)

Comments

[James A Simmons]

Patch landed that fixes just conf-sanity 32 test. Other test will be fixed under LU-18652

[James A Simmons]

One patch landed and one more to go.

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/58115/
Subject: LU-18728 tests: use urandom to really consume ZFS space
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 6153eed3ee180e8695c9e2e3d9ad9db8a6b73ad8

[James A Simmons]

There are 2 patches for this ticket so once one patch lands please don't close.

[Peter Jones]

Bruno

I think that you should be able to find what you need on the test parameters page of the Whamcloud wiki

Peter

[Bruno Faccini]

Gerrit #58115 is an attempt to fix these tests accordingly to Gerrit #57871 for LU-18646.

I am missing (forgot...) the way/tag to only run with ZFS back-end , can I get some hint ? Thx in advance !

 

[Gerrit Updater]

"Bruno Faccini <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/58115
Subject: LU-18728 tests: use urandom to really consume ZFS space
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: ac427f2272cb065b0025b069f44e9b409f4bc32d

[James A Simmons]

The config-sanity test are know about under LU-18652. I disabled those test so they shouldn't be failing anymore. Are you running the latest master? There was another ticket for the 413a timeouts. I haven't seen them but if they still happen then we need to look at the failures. Peter we had a ticket for the 413a timeouts but I closed out the ticket since the failures seemed to go away. Should I reopen that ticket.