[LU-13667] ptlrpc_pinger_main is stuck in endless loop

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.14.0, Lustre 2.12.6
• Labels: llnl
• Severity: 3
• Assignee: Hongchao Zhang
• Reporter: Hongchao Zhang

Description

In ptlrpc_pinger_main, the process of the pingable imports or obd_update_maxusage
could cost long time and be stuck in endless loop because of the negative timeout
returned by pinger_check_timeout

Issue Links

• is related to: LU-14046 lov tgt 0 not cleaned! deathrow=0, lovrc=1

Comments

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/39344/
Subject: LU-13667 ptlrpc: fix endless loop issue
Project: fs/lustre-release
Branch: b2_12
Current Patch Set:
Commit: 95cd26446e16c63b531ed94a844b5f69c8b3730f

[Gerrit Updater]

Minh Diep ([e-mail]) uploaded a new patch: https://review.whamcloud.com/39344
Subject: LU-13667 ptlrpc: fix endless loop issue
Project: fs/lustre-release
Branch: b2_12
Current Patch Set: 1
Commit: 5acd6853b4a64057ce55174a15a93b11d2922eab

[Peter Jones]

Landed for 2.14

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/38915/
Subject: LU-13667 ptlrpc: fix endless loop issue
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 6be2dbb2595121fabceda86c5f7bdcb45e10b320

[Olaf Faaland]

Thank you. This should go into b2_12 after it's merged to master.

[Hongchao Zhang]

Yes, it should be the same issue with this ticket.

[Olaf Faaland]

Please let me know whether you agree my described symptoms match this issue, thanks.

[Olaf Faaland]

I believe we hit this today on a Lustre 2.12.4 client. The pinger was taking 100% of a core. Over 498 seconds, the "next wakeup in" message appeared in the debug log 76,716 times, and the time_to_next_wake started at -41,852 and ended at -42,350 (getting more and more negative with time).

[Gerrit Updater]

Hongchao Zhang ([e-mail]) uploaded a new patch: https://review.whamcloud.com/38915
Subject: LU-13667 ptlrpc: fix endless loop issue
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 9ccad707e88b3b2bc118eac588817737cb9da1c9