[ https://issues.jboss.org/browse/JGRP-1825?page=com.atlassian.jira.plugin.... ] Manuel Dominguez Sarmiento commented on JGRP-1825: -------------------------------------------------- The following articles give a good overview of the current design: http://codespot.net/2014/02/03/hibernate-caching-strategies/ http://clustermania.blogspot.com.ar/2009/07/with-read-write-hibernate-2nd... However, the soft-lock algorithm for the READ-WRITE strategy requires that, for being effective in a cluster: a) Updates to the cluster are replicated synchronously. This is something that the Ehcache-JGroups replication module does somewhat poorly, because even though "synchronous" replication is supported, the implementation does not wait for the RPC acknowledgement - we've patched this some years ago and we maintain our own fork, since Terracotta stopped any this of development of clustering solutions for Ehcache which do not involve their commercial offerings. b) Read-write locks guarding access the the replicated synchronous cache, otherwise transactional consistency is compromised. -- This message was sent by Atlassian JIRA (v6.4.11#64026)