[jboss-jira] [JBoss JIRA] (JBMESSAGING-1918) Failover may take a long time blocking on sucker connection retry

Failover may take a long time blocking on sucker connection retry ----------------------------------------------------------------- Key: JBMESSAGING-1918 URL: https://issues.jboss.org/browse/JBMESSAGING-1918 Project: JBoss Messaging Issue Type: Bug Components: JMS Clustering Affects Versions: 1.4.0.SP3.CP14, 1.4.8.SP6 Reporter: Yong Hao Gao Assignee: Yong Hao Gao Fix For: 1.4.0.SP3.CP15, 1.4.8.SP7 When a node leave the cluster, its failover node will detects the sucker connection broken and try to cleanup and retry. The cleanup and retry method may take a long time to finish as the connection is already broken and the cleanup tries to make remote calls (although they will eventually fail) as part of the cleanup. The methods org.jboss.messaging.core.impl.clusterconnection.ClusterConnectionManager.ConnectionInfo.cleanupConnection() and org.jboss.messaging.core.impl.clusterconnection.ClusterConnectionManager.ConnectionInfo.retryConnection() are all synchronized on ConnectionInfo. On the other hand the node will at the same time performs failover for that node, and this include removing suckers from the node. The method org.jboss.messaging.core.impl.clusterconnection.ClusterConnectionManager.ConnectionInfo.removeSucker(String) synchronizes on ConnectionInfo too. So it's possible that removeSucker() will wait for a long time to get the lock, making the failover process very slow in some situations.