Topology cache updates on stop should be asynchronous ----------------------------------------------------- Key: ISPN-1245 URL: https://issues.jboss.org/browse/ISPN-1245 Project: Infinispan Issue Type: Bug Components: Cache Server Affects Versions: 5.0.0.CR7 Reporter: Galder Zamarreño Assignee: Galder Zamarreño Fix For: 5.0.0.FINAL Topology cache updates when a server stops not only need Flag.FAIL_SILENTLY, Flag.ZERO_LOCK_ACQUISITION_TIMEOUT flags, but also need Force.FORCE_ASYNCHRONOUS to avoid waiting from replies from other nodes. Remember that there's a crashed member view listener running in the coordinator of the cluster that checks whether the view was not updated and whether any old members are still lingering in the view, and updates it accordingly. It could happen that the coordinator has the right view, cos the node on stop updated its view, but not another node's view. This is not too bad cos the client could do a failover if a node is not available and it'd get eventually a new view. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira