[LU-1209] sanity.sh subtest test_133d failed with "samedir_rename_size count error"

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.3.0
• Affects Version/s: Lustre 2.4.1, Lustre 2.6.0
• Labels: yuc2
• Severity: 3
• Assignee: Di Wang (Inactive)
• Reporter: Maloo

Description

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

This issue relates to the following test suite run: https://maloo.whamcloud.com/test_sets/025d7aac-6ce3-11e1-9174-5254004bbbd3.

This subtest is reporting 14% failure over the past 100 runs, so there must be some race or unexpected result causing the failure.

The sub-test test_133d failed with the following error:

samedir_rename_size count error

Info required for matching: sanity 133d

Issue Links

• is duplicated by: LU-1266 2.1.1<->2.2 Test failure on test suite sanity, subtest test_133c
• is related to: LU-1193 test script incompatibility when running server as 2.1 and client as 2.2
• Trackbacks: Developer Work Area Sequoia Blackboard

Comments

[James Nunez (Inactive)]

Hit this on review-ldiskfs https://maloo.whamcloud.com/test_sets/ecfe391a-c41c-11e3-a793-52540035b04c

[Nathaniel Clark]

Hit issue on review-zfs on master (pre 2.6):
https://maloo.whamcloud.com/test_sets/d1c63508-a095-11e3-aa36-52540035b04c

[Jian Yu]

Here is the fix http://review.whamcloud.com/#change,3298

The above patch exists on Lustre b2_4 branch build #67. However, the test still failed:
https://maloo.whamcloud.com/test_sets/596be6ba-6351-11e3-8c76-52540035b04c

[Li Wei (Inactive)]

https://maloo.whamcloud.com/test_sets/ecc68362-a14f-11e2-b1c3-52540035b04c

[Sarah Liu]

Hit this issue again when upgrade from 1.8.9 to 2.4 and then add one new MDT:

https://maloo.whamcloud.com/test_sets/a02cc9b2-9ec5-11e2-975f-52540035b04c

[Jodi Levi (Inactive)]

Please reopen ticket if additional work is needed.

[Di Wang (Inactive)]

Ah, we should get dir size after rename, since rename will change the dir size. Here is the fix http://review.whamcloud.com/#change,3298

[Di Wang (Inactive)]

Hmm, the dir size is 32K, but somehow it record rename under 64K size. it seems dir size is not consistency during the whole process.

[Andreas Dilger]

== sanity test 133d: Verifying rename_stats ====== 19:02:48 (1341626568)
mdt.lustre-MDT0000.rename_stats
mdt.lustre-MDT0000.rename_stats=clear
total: 512 creates in 0.91 seconds: 565.37 creates/second
source rename dir size: 32K
target rename dir size: 4K
mdt.lustre-MDT0000.rename_stats=
rename_stats:
- snapshot_time:  1341626570.213120
- same_dir       
      64KB: { sample:   1, pct: 100, cum_pct: 100 }
/usr/lib64/lustre/tests/sanity.sh: line 7552: [: : integer expression expected
 sanity test_133d: @@@@@@ FAIL: samedir_rename_size error

[Andreas Dilger]

Again: https://maloo.whamcloud.com/test_sets/2d51efd8-c7fe-11e1-ba35-52540035b04c

It is currently reporting a 29% failure rate in Maloo.

[Ian Colle (Inactive)]

Happened again: https://maloo.whamcloud.com/test_sets/63204c18-c4d4-11e1-af06-52540035b04c