[LU-17450] sanity: interop test failures with master+2.15

• Type: Bug
• Resolution: Fixed
• Priority: Critical
• Fix Version/s: Lustre 2.16.0
• Severity: 3
• Assignee: Hongchao Zhang
• Reporter: Maloo

Description

This issue was created by maloo for Andreas Dilger <[e-mail]>

This issue relates to the following test suite run: https://testing.whamcloud.com/test_sets/dc77145c-b7d3-4010-a7a2-f8435f9353ff

Test session details:
clients: https://build.whamcloud.com/job/lustre-reviews/101478 - 4.18.0-477.27.1.el8_8.x86_64
servers: https://build.whamcloud.com/job/lustre-b2_15/81 - 4.18.0-513.9.1.el8_lustre.x86_64

there are a number of sanity interop test failures with 2.15.4 servers

Please review the test failures to determine for each one:

• when did the failure first start happening?
• is this a new test added since 2.15.50 was forked from b2_15?
• should the test be skipped because of an older MDS or OSS version?
• is this a legitimate regression?
• if not a clear case of new test with old server, ensure an LU ticket is open for it with details and add to always_except

Issue Links

• is related to: LU-17525 Unaligned DIO interop with different page sizes fails
• is related to: LU-17597 interop: master/2.15/2.14/2.12 sanity test_56x: migrate failed rc = 22
• is related to: LU-13805 i/o path: Unaligned direct i/o
• is related to: LU-17216 enable_health_write, health_check improvements
• is related to: LU-14361 Add support for statahead pattern with regularized file name
• is related to: LU-16194 Define negative PFL extent start/end as invalid

Comments

[Peter Jones]

Merged for 2.16

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/54962/
Subject: LU-17450 test: disable test 56x 56xa 56xb in sanity
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: bfcab56c5f35ae33c42325326f9141d9c48412b1

[Gerrit Updater]

"Hongchao Zhang <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/54962
Subject: LU-17450 test: disable test 56x 56xa 56xb in sanity
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: d763aa7873e9160f3e406a4066dbc114bd0320d8

[Hongchao Zhang]

the interop failure of test_56x, test_56xa, test_56xb in sanity could be fixed by the patch
https://review.whamcloud.com/53997/ in LU-17525

[Peter Jones]

Hongchao

Can you please investigate what needs to be done to avoid this test failure?

Thanks

Peter

[Andreas Dilger]

I've requested patches to add interop checks for the identified source patches, but not sure why test_56x is failing. Those tests look related to lfs_migrate, but needs some more investigation why it is failing.

[Andreas Dilger]

test_56x, test_56xa, test_56xc - not sure (they are old tests)
test_65p - from LU-16194 lod: define negative extent offset as invalid
test_70a - from LU-17216 ofd: make enable_health_write tunable, requested interop check there
test_119h, test_119i, test_398d, test_398o - from LU-13805 clio: bounce buffer for unaligned DIO
test_123g, test_123h, test_123i - from LU-14361 statahead: Add test for statahead advise

[Andreas Dilger]

Make this a blocker for 2.16.0 since we can't release it until we are sure it is not introducing any interop issues, which would be much harder to fix afterward.

It looks at a minimum that there is something wrong with file migrate, but I haven't looked at all of the failures in this interop session yet. There is a small chance that one or two subtest failures relate to the patch that was being tested (which is why I requested interop testing in the first place), but many of the failures have been present for weeks and are not isolated to a single subtest.