[ https://issues.jboss.org/browse/ISPN-1855?page=com.atlassian.jira.plugin.... ] Dan Berindei commented on ISPN-1855: ------------------------------------ With the ISPN-2632 fix we send the cache topology id to the client instead of the JGroups view id, so accessing a replicated (or even local-mode) cache won't break future updates. But the problem remains that the client only keeps one CH for each HotRod cache manager, and the server has a different CH for each cache. So if a node joins and a client's first request is for a replicated cache, the topology id on the client is incremented (by 2) and the next request for a distributed cache won't receive any topology update. The proper solution here is to keep one CH per cache on the client, as well. Note that sending a CH for replicated caches isn't necessarily a good idea - all requests would go to the coordinator, because that's the primary owner of all the keys in replicated mode. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira