Are you using 5.6, this looks a bit like https://issues.apache.org/jira/browse/AMQ-3706
The test case attached to that jira may provide a good starting point for a junit test case for your scenario, that could help reproduce http://svn.apache.org/viewvc/activemq/trunk/activemq-core/src/test/java/org/apache/activemq/transport/failover/FailoverComplexClusterTest.java?view=markup On 22 June 2012 16:32, Pippo <shy.if...@optier.com> wrote: > Okay, > > So I was able to achieve that by configuring the third broker like this: > > /<networkConnectors> > <networkConnector name="nc-failover" > uri="static:(tcp://amq1:61616,tcp://amq2:61616,tcp://amq3:61616)" > decreaseNetworkConsumerPriority="true" > networkTTL="1" > <b>duplex="true" />* > </networkConnectors> > : > <transportConnectors> > <transportConnector name="openwire" > uri="tcp://0.0.0.0:61616" > updateClusterClients="true" > rebalanceClusterClients="true" > updateClusterClientsOnRemove="true" /> > </transportConnectors>/ > > When I bring up the third broker, the clients rebalance themselves among the > several nodes. > Now, I want to update the configuration of the first two brokers. So I take > down one of them and there are times when I see that some of the clients > reports on a transport error and doesn't reconnect to the remaining brokers. > > Does this sound like a bug in ActiveMQ? > > Thanks. > > > -- > View this message in context: > http://activemq.2283324.n4.nabble.com/Adding-nodes-to-Cluster-without-restarting-tp4653478p4653491.html > Sent from the ActiveMQ - User mailing list archive at Nabble.com. -- http://fusesource.com http://blog.garytully.com
