Red Hat Bugzilla – Bug 485255
Cluster: message producers don't always seem to failover cleanly.
Last modified: 2011-01-25 16:17:19 EST
Description of problem:
Message producers don't always seem to failover cleanly. Error
2009-02-09 18:23:40,859 ERROR [IoReceiver -
(AMQConnectionDelegate_0_10.java:234) - error during failover
org.apache.qpid.transport.SessionException: timed out waiting for
session to become open (state=DETACHED)
Rafi has fixed this error condition. Now you should get a more clean error message. If you can provide me with the test case I could test it as well.
Hi on which version do you see this issue? How often How to reproduce this? Raising NEEDINFO
no info yet ..
I assume this error was reported on the 1.1 release as thats the version mentioned in BZ.
I can't really recall the exact circumstances that triggered this issue, other than to say that this issue is more likely to happen when you recycle the broker the client is connected to repeatedly while the producer/consumer pair is in flight.
A primitive version of current failover tests were used for this. Therefore it's safe to say that the current failover test used by you and the one in testkit is adequate in testing this issue.
Technical note added. If any revisions are required, please edit the "Technical Notes" field
accordingly. All revisions will be proofread by the Engineering Content Services team.
Cause: Message producers don't always seem to failover cleanly to another member in the cluster.
Consequence: The client appears to hang and then eventually times out without logging/notifying the proper error message.
Fix: The error condition is handled and notified promptly.
Result: The client does not hang anymore, instead it prints the error condition clearly.