Details
-
Bug
-
Resolution: Done
-
Major
-
7.1.2.Final (EAP)
-
None
Description
Hi,
I hit an issue with HornetQ core bridges. I tried test scenario with 4 nodes where:
1) there are two clusters A and B (each contains two nodes)
2) there is hornetq core bridge configured from cluster A to cluster B (A1 -> B1, A2 -> B2)
During sending messages (30KB ByteMessages) to A1 and reading them from B2, node A2 was killed and started a few times. In the end there were duplicated messages. Problem was in low value of id-cache-size attribute which is 2000 by default and confirmation-window-size which is for bridges (1024 * 1024) bytes.
Can be default value of id-cache-size increased so we can avoid this situation? This is potentially dangerous and hard for admins to realize before the problem occurs.