There doesn't seem to be a Users list of any kind, so I'll make my plea here.
We're running:
Message Queue 5.1
Oracle
Version: 5.1 (Build 9-b)
Compile: July 29 2014 1229
We have had a 2 leg cluster for years, it's simple and straightforward.
Currently we're trying to add a 3rd leg to the cluster. But the 3rd instance can not connect to the other two.
It's properly connecting to the DB, the other 2 can see the 3rd leg, in that they're aware of its existence and when it seems to "go away". But we get errors of the 3rd leg trying to connect directly to the other individual instances in the cluster.
[#|2020-01-09T00:16:26.324-0800|WARNING|5.1|imq.log.Logger|_ThreadID=30;_ThreadName=Thread-jms[0];|WARNING
[B2191]: Timeout(180) in waiting for cluster connection to
mq://10.40.60.21:7777/?instName=???&brokerID=brokerngsharemail1&brokerSessionUID=null&ha=true[/10.40.60.21:34650]
[#|2020-01-09T00:16:26.357-0800|INFO|5.1|imq.log.Logger|_ThreadID=32;_ThreadName=Thread-jms[2];|[B1066]:
Closing: ???@10.40.60.37:0->jms:0 because "Destroying old
connection with same connectionUID 9124134005846419200 - reconnect is
happening before
connection was reaped". Count: service=2 broker=2
[#|2020-01-09T00:16:26.364-0800|INFO|5.1|imq.log.Logger|_ThreadID=32;_ThreadName=Thread-jms[2];|[B1066]:
Closing: ???@10.40.60.37:0->jms:0 because "[B0063]: Initialization
Failure ". Count: service=1 broker=1
This should be straight forward, we've tested all of the network parts. We can ping, and telnet, and nc from each leg to the other. We don't think it's a firewall issue (if it was, what could it be?)
The configuration is simple, MQ isn't difficult to configure, it pretty much "just works" for us and in the past.
It's not using UDP is it? (That just occurred to me, we haven't tested that, but, honestly, I don't think it would be using UDP...)
The instances are running independently, they're not embedded in an app server. Just stand alone instances.
Any insight in to this would be helpful.
Thanks.
Regards,
Will Hartung
