[LU-5748] sanity-lfsck test_9b: unexpected status

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.7.0
• Type: Bug
• Priority: Major
• Reporter: Maloo
• Assignee: nasf (Inactive)
• Resolution: Duplicate
• Severity: 3

Description

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

This issue relates to the following test suite run: https://testing.hpdd.intel.com/test_sets/ed4f6082-53f1-11e4-a9db-5254006e85c2.

The sub-test test_9b failed with the following error:

(5) unexpected status

Please provide additional information about the failure here.

Info required for matching: sanity-lfsck 9b

Comments

[nasf (Inactive)]

According to the log on MDT, the LFSCK assistant thread triggered the error stub (#define OBD_FAIL_LFSCK_NO_DOUBLESCAN 0x160c) as expected, and began to handle the exiting, but the test timeout before the assistant thread exit.

00100000:10000000:0.0:1413319551.677715:0:28817:0:(lfsck_engine.c:1568:lfsck_assistant_engine()) lustre-MDT0000-osd: LFSCK assistant phase2 scan start
00100000:02000000:0.0:1413319551.677721:0:28817:0:(libcfs_fail.h:96:cfs_fail_check_set()) *** cfs_fail_loc=160c, val=0***
00100000:10000000:0.0:1413319551.679212:0:28817:0:(lfsck_engine.c:1662:lfsck_assistant_engine()) lustre-MDT0000-osd: LFSCK assistant unknown status: rc = 0
00080000:00080000:0.0:1413319551.679229:0:28817:0:(osd_handler.c:541:osd_sync()) syncing OSD osd-zfs
...

So there are two possible cases:
1) The test_9b should wait more long time for the LFSCK to commit async updating.
2) Something wrong inside the dt_sync() and caused the LFSCK assistant thread could not exit in time.

Since there are no sync related failures in the logs, and there are some known performance issues for zfs-based backend, I more suspect that we should increase the lfsck tests timeout. There is another ticket for that LU-5301.

[nasf (Inactive)]

It is another instance of LU-5301.