[LU-5571] Test failure sanity-lfsck test_13: (2) unexpected status

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.7.0
• Type: Bug
• Priority: Critical
• Reporter: Maloo
• Assignee: nasf (Inactive)
• Resolution: Cannot Reproduce
• Severity: 3

Description

This issue was created by maloo for Nathaniel Clark <[e-mail]>

This issue relates to the following test suite run:
https://testing.hpdd.intel.com/test_sets/c07fa6c6-189a-11e4-bd79-5254006e85c2
https://testing.hpdd.intel.com/test_sets/fe33929a-2fe0-11e4-957a-5254006e85c2
https://testing.hpdd.intel.com/test_sets/fe22e280-3146-11e4-b503-5254006e85c2
https://testing.hpdd.intel.com/test_sets/199c6a08-31d9-11e4-a833-5254006e85c2

The sub-test test_13 failed with the following error:

(2) unexpected status

Info required for matching: sanity-lfsck 13

Issue Links

• is related to: LU-3534 async update cross-MDTs

Comments

[Jodi Levi (Inactive)]

Fan Yong,
Can you comment on this one please?
Thank you!

[nasf (Inactive)]

The zfs-based backend seems some slow. I have changed the test scripts (in the patch http://review.whamcloud.com/#/c/10987/ set 16) to wait more long time for the LFSCK status changing. Let's see what will happen after such patch applied.

[Jodi Levi (Inactive)]

Nathaniel,
Could you run the test with this patch applied to see if it corrects the issue?
Thank you!

[Nathaniel Clark]

I cannot reproduce with 10987 applied and I don't see any failures on builds that include the patch (10987).

[nasf (Inactive)]

Close it since it is cannot be reproduced after the patch applied.

[Andreas Dilger]

There are a large number of these failures being generated by the LU-3534 patch series, in particular:

http://review.whamcloud.com/10794
http://review.whamcloud.com/10939

but I'm tracking those under LU-3534 instead of this ticket, which may not be related.

[nasf (Inactive)]

The sanity-lfsck failures for DNE are not the same. The original test_13 failed because the low layer iteration was too slow to complete the LFSCK in time. The general behavior is that the final status was "scanning-phase2", but not the expected "completed".

The failures for the DNE patches were that the test scripts cannot find the specified lproc interface "error: get_param: mdd/lustre-MDT0000/lfsck_layout: Found no match". I am still not sure what caused such failure, but it should not be related with the iteration performance.