Map<String,Object> connectionParams1 = new HashMap<String,Object>();
connectionParams1.put(TransportConstants.HOST_PROP_NAME, 127.0.0.1);
connectionParams1.put(TransportConstants.PORT_PROP_NAME, 5445);
TransportConfiguration transportConfiguration1 = new TransportConfiguration(NettyConnectorFactory.class.getName(), connectionParams1);

Map<String,Object> connectionParams2 = new HashMap<String,Object>();
connectionParams2.put(TransportConstants.HOST_PROP_NAME, 127.0.0.1);
connectionParams2.put(TransportConstants.PORT_PROP_NAME, 5447);
TransportConfiguration transportConfiguration2 = new TransportConfiguration(NettyConnectorFactory.class.getName(), connectionParams2);

serverLocator = HornetQClient.createServerLocatorWithHA(transportConfiguration1, transportConfiguration2);
serverLocator.setRetryInterval(1000);
serverLocator.setRetryIntervalMultiplier(1.0);
serverLocator.setReconnectAttempts(1);

clientSessionFactory = serverLocator.createSessionFactory();
clientSession = clientSessionFactory.createSession(userName, password, false, true, true, true, 1);
clientSession.start();

Unfortunately above code didn’t worked on Windows, but after putting below patch it worked on Windows for failover. But its not working on Linux, when both servers and client run on Linux.

Corresponding code works for JMS API both on windows and Linux.

for(int j=0;j<4;j++) {
if(clientSessionFactory != null)
  clientSessionFactory.close();
clientSessionFactory = serverLocator.createSessionFactory();
}

Its working fine now. Fix is to remove second entry transportConfiguration2 which points to backup server. Very strange, I dont know why it works on Windows but not on AWS Linux node.

Hi,

Conclusion:

Real culprit was serverLocator.setReconnectAttempts(1); I increased it to 10, transportConfiguration2 is fine.

Thnx