-
Bug
-
Resolution: Done
-
Blocker
-
None
In a rare race, the journals may get out of order after compacting leading to lose a transaction, if server is killed right after the out of order data is recorded.
It's fairly rare to reproduce it. It was reproduced after WFIP-119, and I couldn't reproduce it in 7.1.x but I could verify the issue also exists in 7.1.x.
So, I'm raising a fix before it's reported.
- is incorporated by
-
JBEAP-14987 [GSS](7.1.z) Upgrade ActiveMQ Artemis from 1.5.5.jbossorg-013 to 1.5.5.jbossorg-014
- Closed
- relates to
-
WFWIP-119 [Artemis 2.x upgrade] Lost message in replicated HA scenario with MDB
- Resolved