I have 3 servers (A,B,C) set up in a chained cluster. Connections to each other are defined explicitly.
A connects to B.
B connects to A and C.
C connects to B.
The cluster connections on all three are defined to cluster the address jms.
Server A cluster config looks like this:
I have no queues or topics defined on Server B. On server A and C I have a jms queue defined:
I can start up A, B, C and connect a producer and consumer and everything works well. Then I kill server B, and start it back up. I get an exception (pasted below) and no more messages are delivered. If I then stop C and start it again, and reconnect the consumer, messages flow again, but any sent during that down period are not consumed.
I also see this message every five seconds on server B with the command id incrementing :
[Old I/O client worker (channelId: 20259911, /(server b's ip):3863 => (server c's host name/server c's ip:5445)] 15:32:44,255 WARNING [org.hornetq.core.protocol
.core.impl.ChannelImpl] Can't find packet to clear: last received command id 5
8 first stored command id 0