[LU-18293] sanity-quota test_1a, 1b, 1c: FAIL: user write success, but expect EDQUOT

• Type: Bug
• Resolution: Unresolved
• Priority: Minor
• Fix Version/s: Lustre 2.16.0, Lustre 2.17.0
• Affects Version/s: Lustre 2.16.0
• Labels: zfs
• Severity: 3
• Assignee: Hongchao Zhang
• Reporter: Jian Yu

Description

sanity-quota test_1a, 1b, and 1c failed on 2.16.0 RC1 full-zfs-part-2 test session:
https://testing.whamcloud.com/test_sets/0d7fdd5a-b404-4b82-bd32-aaf538cee475

Disk quotas for grp quota_usr (gid 60000):
      Filesystem  kbytes   quota   limit   grace   files   quota   limit   grace
     /mnt/lustre   10245*      0   10240       -       1       0       0       -
  lustre-MDT0000       2       -       0       -       1       -       0       -
  lustre-OST0000       0       -       0       -       -       -       -       -
  lustre-OST0001       0       -       0       -       -       -       -       -
  lustre-OST0002       0       -       0       -       -       -       -       -
  lustre-OST0003   10244       -       0       -       -       -       -       -
  lustre-OST0004       0       -       0       -       -       -       -       -
  lustre-OST0005       0       -       0       -       -       -       -       -
  lustre-OST0006       0       -       0       -       -       -       -       -
Total allocated inode limit: 0, total allocated block limit: 0
Files for group (quota_usr), count=1:
  File: /mnt/lustre/d1a.sanity-quota/f1a.sanity-quota-1
  Size: 11534336  	Blocks: 20487      IO Block: 4194304 regular file
Device: 2c54f966h/743766374d	Inode: 144117285664063501  Links: 1
Access: (0644/-rw-r--r--)  Uid: (60000/quota_usr)   Gid: (60000/quota_usr)
Access: 2024-10-01 10:01:26.000000000 +0000
Modify: 2024-10-01 10:01:36.000000000 +0000
Change: 2024-10-01 10:01:36.000000000 +0000
 Birth: 2024-10-01 10:01:26.000000000 +0000
 sanity-quota test_1a: @@@@@@ FAIL: user write success, but expect EDQUOT 

Issue Links

• is related to: LU-15299 sanity-quota test_71a: FAIL: user write failure, but expect success

Comments

[Alex Zhuravlev]

with this patch landed sanity-quota/71a doesn't pass locally, I did a bisection:

COMMIT          TESTED  PASSED  FAILED          COMMIT DESCRIPTION
2f96622a52      5       0       5       BAD     LU-18293 test: use direct IO for quota test
1811e77562      5       5       0       GOOD    LU-16134 utils: un-deprecate 'lctl set_param -F'

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/58533/
Subject: LU-18293 test: use direct IO for quota test
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 2f96622a522924c7b9e26fd2dfdf34eb6227b7a6

[Gerrit Updater]

"Hongchao Zhang <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/58533
Subject: LU-18293 test: use direct IO for quota test
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 43184902d29ed8ed053a5da38e70781398b65f19

[Gerrit Updater]

"Jian Yu <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/57078
Subject: LU-18293 test: check to sync quota setting
Project: fs/lustre-release
Branch: b2_15
Current Patch Set: 1
Commit: a69b054a22e1ae97509b659b086338b5286241fc

[Peter Jones]

Merged for 2.16

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/56628/
Subject: LU-18293 test: check to sync quota setting
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: a9caa85ad2e344ff7ea2a2f33367290600439a13

[Jian Yu]

This failure occurred consistently on the following test sessions:
lustre-master-el8.10-x86_64-full-zfs-part-2
lustre-master-el8.9-x86_64-full-zfs-part-2
lustre-master-el8.8-x86_64-full-zfs-part-2
lustre-master-el9.3-x86_64-full-zfs-part-2
lustre-master-el9.3-x86_64-full-dne-zfs-part-2

[Gerrit Updater]

"Hongchao Zhang <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/56628
Subject: LU-18293 test: check to sync quota setting
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 915a33ead2e15fee1ebae296d964335b53a0278f

[Hongchao Zhang]

As per the logs, this issue could be caused by the out-of-sync of the global quota setting between QMT to QSD, and cause
the new quota setting not to be transferred to QSD, then the corresponding quota ID is not enforced on the QSD.

[Li Xi]

hongchao.zhang Would you please take a look?