[SERVER-56833] Invariant failure in WiredTigerIndexUnique::_keyExists Created: 11/May/21  Updated: 02/Jul/21  Resolved: 02/Jul/21

Status: Closed
Project: Core Server
Component/s: Storage
Affects Version/s: 4.4.5
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Dmitry Agranat Assignee: Pavithra Vetriselvan
Resolution: Duplicate Votes: 0
Labels: Atlas_Failure_Analysis
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
Related
Operating System: ALL
Sprint: Repl 2021-06-14, Repl 2021-06-28, Repl 2021-07-12
Participants:
Linked BF Score: 15

 Description   

{"t":{"$date":"2021-05-03T21:53:13.408+00:00"},"s":"F",  "c":"-",        "id":23079,   "ctx":"ReplWriterWorker-13","msg":"Invariant failure","attr":{"expr":"client->shouldKillSystemOperation(lk)","file":"src/mongo/db/storage/wiredtiger/wiredtiger_prepare_conflict.h","line":87}}



 Comments   
Comment by Pavithra Vetriselvan [ 02/Jul/21 ]

I'm fairly confident that SERVER-58183 is the issue we're seeing here, so I'm going to mark this as a duplicate of the 4.4 backport of that ticket. Please see this comment on BF-20067 for an in-depth analysis.

Comment by Pavithra Vetriselvan [ 30/Jun/21 ]

Based on the stack trace, this is probably a dupe of BF-20067 as Louis mentioned.

We unfortunately don't have level 4 debug logging for transactions, so we can't confirm that we restarted a transaction while applying the prepare. However, from the stack trace we can clearly see that we're hitting a prepare conflict from _applyOperationsForTransaction, which is the same symptom as BF-20067.

Generated at Thu Feb 08 05:40:19 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.