[LU-6016] sanity-lfsck test_20a: FAIL: (1) Fail to start MDS!

• Type: Bug
• Resolution: Fixed
• Priority: Blocker
• Fix Version/s: Lustre 2.5.4
• Affects Version/s: Lustre 2.5.4
• Environment: Lustre build: https://build.hpdd.intel.com/job/lustre-b2_5/106/
• Severity: 3
• Assignee: Jian Yu
• Reporter: Jian Yu

Description

sanity-lfsck test 20a failed as follows:

start mds1
CMD: shadow-50vm7 mkdir -p /mnt/mds1
CMD: shadow-50vm7 test -b /dev/lvm-Role_MDS/P1
CMD: shadow-50vm7 mkdir -p /mnt/mds1; mount -t lustre -o user_xattr  		                   /dev/lvm-Role_MDS/P1 /mnt/mds1
shadow-50vm7: mount.lustre: according to /etc/mtab /dev/mapper/lvm--Role_MDS-P1 is already mounted on /mnt/mds1
 sanity-lfsck test_20a: @@@@@@ FAIL: (1) Fail to start MDS! 

Maloo report: https://testing.hpdd.intel.com/test_sets/c0e13104-8071-11e4-a434-5254006e85c2

Issue Links

• is related to: LU-7682 Interop 2.7.1<->master - sanity-benchmark: no subtests ran

Comments

[Jian Yu]

Patch was landed on Lustre b2_5 branch.

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch http://review.whamcloud.com/13025/
Subject: LU-6016 tests: do not need start MDS in sanity-lfsck test 20a
Project: fs/lustre-release
Branch: b2_5
Current Patch Set:
Commit: e835226b17309ec21fd7b46cf397e5fd557049bd

[Jian Yu]

This is blocking review testing on Lustre b2_5 patches.

[Gerrit Updater]

Jian Yu ([e-mail]) uploaded a new patch: http://review.whamcloud.com/13025
Subject: LU-6016 tests: do not need start MDS in sanity-lfsck test 20a
Project: fs/lustre-release
Branch: b2_5
Current Patch Set: 1
Commit: f1ad071041152605ead72ca4ca432c1231249e4b

[Jian Yu]

The root cause is that patch http://review.whamcloud.com/12740 was submitted without considering patch http://review.whamcloud.com/10818, which removed formatall and setupall from lfsck_prep(). So, while running review test session, test 20a passed with reformatting the filesystem. After the two patches were landed on Lustre b2_5 branch, test 20a failed because formatall and setupall were removed.

I'm creating a patch to fix test 20a to remove starting $SINGLEMDS.

[Jian Yu]

This is a regression failure introduced by the patch http://review.whamcloud.com/12740 for LU-4958 in Lustre b2_5 build #106.

It looked like $SINGLEMDS was already mounted before running test 20a. However, it's strange that in the patch review test session, test 20a passed.