[LU-5063] sanity-quota test_7b test_12a test_13: dd failed

• Type: Bug
• Resolution: Unresolved
• Priority: Minor
• Severity: 3
• Assignee: WC Triage
• Reporter: Maloo

Description

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

This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/5577099a-dbf1-11e3-8f5e-52540035b04c.

The sub-test test_12a failed with the following error:

sanity-quota test_12a: @@@@@@ FAIL: dd failed
Trace dump:
= /usr/lib64/lustre/tests/test-framework.sh:4492:error_noexit()
= /usr/lib64/lustre/tests/test-framework.sh:4523:error()
= /usr/lib64/lustre/tests/sanity-quota.sh:154:quota_error()
= /usr/lib64/lustre/tests/sanity-quota.sh:1426:test_12a()
= /usr/lib64/lustre/tests/test-framework.sh:4784:run_one()
= /usr/lib64/lustre/tests/test-framework.sh:4819:run_one_logged()
= /usr/lib64/lustre/tests/test-framework.sh:4672:run_test()
= /usr/lib64/lustre/tests/sanity-quota.sh:1444:main()
Dumping lctl log to /logdir/test_logs/2014-05-14/lustre-reviews-el6-x86_64-review-dne-part-2-2_9_1_23894_-70192678855860-141811/sanity-quota.test_12a.*.1400108600.log
CMD: shadow-16vm10.shadow.whamcloud.com,shadow-16vm11,shadow-16vm12,shadow-16vm8,shadow-16vm9 /usr/sbin/lctl dk > /logdir/test_logs/2014-05-14/lustre-reviews-el6-x86_64-review-dne-part-2-2_9_1_23894_-70192678855860-141811/sanity-quota.test_12a.debug_log.$(hostname -s).1400108600.log;
dmesg > /logdir/test_logs/2014-05-14/lustre-reviews-el6-x86_64-review-dne-part-2-2_9_1_23894_-70192678855860-141811/sanity-quota.test_12a.dmesg.$(hostname -s).1400108600.log
Delete files...

Info required for matching: sanity-quota 12a

Issue Links

• duplicates: LU-5082 Test failure sanity-quota test_13: dd failed
• is related to: LU-5081 Test failure on test suite sanity-quota, subtest test_7b

Comments

[Niu Yawei (Inactive)]

Not sure why the used space for quota_usr isn't cleared, that's why all these tests failed.

'Used space(5112) for user quota_usr isn't 0.' 

test_7b/7c failed at the very beginning when checking environment.

dd: writing `/mnt/lustre/d12a.sanity-quota/f12a.sanity-quota-0': Disk quota exceeded

test_12a dd failed for EDQUOT, because the user has already used some space before testing.

Each quota test will remove all files on test exit:

cleanup_quota_test() {
        trap 0
        echo "Delete files..."
        rm -rf $DIR/$tdir
        echo "Wait for unlink objects finished..."
        wait_delete_completed
        sync_all_data || true
}

So I guess some commit broke the 'wait_delete_completed'? (Accessing maloo is extremely slow here, it's hard to tell when such failure was introduced)

[Andreas Dilger]

This isn't the first failure in this test run. test_7b and test_7c also failed before test_12a.

[Jodi Levi (Inactive)]

Niu,
Can you comment on this one?
Thank you!