[LU-4541] Interop 2.5<->2.6 failure on test suite conf-sanity test_45

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.6.0
• Type: Bug
• Priority: Minor
• Reporter: Maloo
• Assignee: WC Triage
• Resolution: Duplicate

Environment

server: lustre-master build #1837
client: b2_5

• Severity: 3

Description

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

This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/71605d20-82b2-11e3-b715-52540035b04c.

The sub-test test_45 failed with the following error:

test failed to respond and timed out

Client console shows:

06:17:32:Lustre: DEBUG MARKER: == conf-sanity test 45: long unlink handling in ptlrpcd == 04:44:21 (1390308261)
06:17:36:Lustre: DEBUG MARKER: mkdir -p /mnt/lustre
06:17:39:Lustre: DEBUG MARKER: mount -t lustre -o user_xattr,flock client-6vm3@tcp:/lustre /mnt/lustre
06:17:39:LustreError: 4124:0:(obd_config.c:1341:class_process_proc_param()) lustre-client-ffff88007cbde800: unknown param some_wrong_param=10
06:17:40:LustreError: 11-0: lustre-MDT0000-mdc-ffff88007cbde800: Communicating with 10.10.4.39@tcp, operation mds_connect failed with -11.
06:17:43:LustreError: 11-0: lustre-MDT0000-mdc-ffff88007cbde800: Communicating with 10.10.4.39@tcp, operation mds_connect failed with -11.

Not sure if this is a dup of LU-2746 and after this error, both the client and OST reboot.

Comments

[Andreas Dilger]

This might be caused by TEI-1403. The test_45 code was running fine, and then within a few seconds of the test starting it was declared a timeout and stacks dumped and rebooted. There is no indication in the logs that the test failed or the node crashed.

[Andreas Dilger]

This "unexpected timeout" problem was also hit on LU-4549 (now TEI-1443), but even before the test was running for 3600s.