[LU-8635] interop: sanity test_205: FAIL: old jobstats not expired

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.10.0
• Severity: 3
• Assignee: Emoly Liu
• Reporter: Emoly Liu

Description

It was reported that

== sanity test 205: Verify job stats == 08:07:42 (1470816462)
Waiting 90 secs for update
Updated after 5s: wanted 'nodelocal' got 'nodelocal'
Registered as changelog user cl4
mdt.lustre-MDT0000.job_cleanup_interval=5
jobid_name=id.205.mkdir.8313
Test: mkdir /mnt/lustre/d205.sanity
Using JobID environment variable nodelocal=id.205.mkdir.8313
jobid_name=id.205.rmdir.3433
Test: rmdir /mnt/lustre/d205.sanity
Using JobID environment variable nodelocal=id.205.rmdir.3433
jobid_name=id.205.mknod.3052
Test: mknod /mnt/lustre/f205.sanity c 1 3
Using JobID environment variable nodelocal=id.205.mknod.3052
jobid_name=id.205.rm.17794
Test: rm -f /mnt/lustre/f205.sanity
Using JobID environment variable nodelocal=id.205.rm.17794
jobid_name=id.205.lfs.4388
Test: /usr/bin/lfs setstripe -i 0 -c 1 /mnt/lustre/f205.sanity
Using JobID environment variable nodelocal=id.205.lfs.4388
jobid_name=id.205.touch.32613
Test: touch /mnt/lustre/f205.sanity
Using JobID environment variable nodelocal=id.205.touch.32613
jobid_name=id.205.dd.18785
Test: dd if=/dev/zero of=/mnt/lustre/f205.sanity bs=1M count=1 oflag=sync
Using JobID environment variable nodelocal=id.205.dd.18785
1+0 records in
1+0 records out
1048576 bytes (1.0 MB) copied, 0.0650743 s, 16.1 MB/s
jobid_name=id.205.dd.2526
Test: dd if=/mnt/lustre/f205.sanity of=/dev/null bs=1M count=1 iflag=direct
Using JobID environment variable nodelocal=id.205.dd.2526
1+0 records in
1+0 records out
1048576 bytes (1.0 MB) copied, 0.0263004 s, 39.9 MB/s
jobid_name=id.205.truncate.26370
Test: /usr/lib64/lustre/tests/truncate /mnt/lustre/f205.sanity 0
Using JobID environment variable nodelocal=id.205.truncate.26370
jobid_name=id.205.mv.14305
Test: mv -f /mnt/lustre/f205.sanity /mnt/lustre/d205.sanity.rename
Using JobID environment variable nodelocal=id.205.mv.14305
sleep 2 for expiry
jobid_name=id.205.mkdir.9284
Test: mkdir /mnt/lustre/d205.sanity.expire
Using JobID environment variable nodelocal=id.205.mkdir.9284
 sanity test_205: @@@@@@ FAIL: old jobstats not expired 
  Trace dump:
  = /usr/lib64/lustre/tests/test-framework.sh:4853:error()
  = /usr/lib64/lustre/tests/sanity.sh:11896:test_205()
  = /usr/lib64/lustre/tests/test-framework.sh:5113:run_one()
  = /usr/lib64/lustre/tests/test-framework.sh:5151:run_one_logged()
  = /usr/lib64/lustre/tests/test-framework.sh:4955:run_test()
  = /usr/lib64/lustre/tests/sanity.sh:11918:main()
Dumping lctl log to /tmp/test_logs/1470816456/sanity.test_205.*.1470816477.log
Resetting fail_loc on all nodes...done.
mdt.lustre-MDT0000.job_cleanup_interval=600
Waiting 90 secs for update
Updated after 10s: wanted 'procname_uid' got 'procname_uid'
lustre-MDT0000: Deregistered changelog user 'cl4'
FAIL 205 (26s)
== sanity test complete, duration 32 sec == 08:08:08 (1470816488)
sanity: FAIL: test_205 old jobstats not expired
debug=super ioctl neterror warning dlmtrace error emerg ha rpctrace vfstrace config console lfsck

Note that LU-6920: http://review.whamcloud.com/16753/ landed but still regression is seen.

I will upstream the patch http://review.whamcloud.com/#/c/22404/ to master.