[SERVER-13260] a race condition between the producer thread and stopReplicationAndFlushBuffer() can cause excessively long rollbacks Created: 18/Mar/14  Updated: 14/Apr/16  Resolved: 26/Aug/15

Status: Closed
Project: Core Server
Component/s: Replication
Affects Version/s: 2.6.0-rc1
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Matt Dannenberg Assignee: Benety Goh
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
Operating System: ALL
Participants:

 Description   

This race allows for a stale op to end up in the _buffer in bgsync.cpp, which can lead to excessively long rollbacks (since the time difference is negative and as a result ignores the 30min limit) when the primary becomes secondary again and the op gets put in the oplog



 Comments   
Comment by Eric Milkie [ 16/Jul/15 ]

Possibly gone away after the refactoring, and the fix for SERVER-18994

Generated at Thu Feb 08 03:31:09 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.