[LU-14738] recovery-small: FAIL: remove sub-test dirs failed

• Type: Bug
• Resolution: Duplicate
• Priority: Minor
• Severity: 3
• Assignee: WC Triage
• Reporter: Alex Zhuravlev

Description

== recovery-small test complete, duration 238 sec ==================================================== 20:17:10 (1623010630)
Lustre: DEBUG MARKER: == recovery-small test complete, duration 238 sec ==================================================== 20:17:10 (1623010630)
LustreError: 167-0: lustre-MDT0001-osp-MDT0000: This client was evicted by lustre-MDT0001; in progress operations using this service will fail.
rm: cannot remove '/mnt/lustre/d110h.recovery-small/target_dir/tgt_file': Input/output error
 recovery-small test_148: @@@@@@ FAIL: remove sub-test dirs failed 
Lustre: DEBUG MARKER: recovery-small test_148: @@@@@@ FAIL: remove sub-test dirs failed
  Trace dump:
  = ./../tests/test-framework.sh:6166:error()
  = ./../tests/test-framework.sh:5650:check_and_cleanup_lustre()
  = recovery-small.sh:3158:main()

the root cause is that one MDS got evicted from another one and subsequent RPC (which is part of MDS_REINT(rmdir) processing) observes eviction, returning -EIO.

Issue Links

• is related to: LU-9602 recovery-random-scale test_fail_client_mds: PASS but marked as FAIL

Comments

[Alex Zhuravlev]

a dup of LU-14938, same problem in a different place

[Gerrit Updater]

Alex Zhuravlev ([e-mail]) uploaded a new patch: https://review.whamcloud.com/43935
Subject: LU-14738 tests: recovery-small/110k should wait for MDS
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: b5080c27fb4e4f1805a3e4331f15ee95872e8a5f