[LU-10212] operation ost_read to node [*] failed: rc = -116

• Type: Bug
• Resolution: Fixed
• Priority: Critical
• Fix Version/s: Lustre 2.11.0, Lustre 2.10.4
• Affects Version/s: Lustre 2.11.0
• Labels: patch
• Severity: 3
• Assignee: Alexander Boyko
• Reporter: Alexander Boyko

Description

1504054839 2017 Aug 30 03:00:39 m10029 kern info kernel Lustre: 
lustre02-MDT0005-mdc-ffff880079922000: Connection restored to 
10.50.33.92@o2ib (at 10.50.33.92@o2ib)
1504063447 2017 Aug 30 05:24:07 m10029 kern err kernel LustreError: 
11-0: lustre01-OST0023-osc-ffff88085f97b000: operation ost_read to node 
10.50.32.72@o2ib failed: rc = -116

Based on logs, there were identical locks from the client. And the client canceled one of them because of two, and sent io. The read request came to server with lock handle which was canceled. And the prolong function didn`t check the resource lock tree for this case. So client got ESTALE error.