[infinispan-issues] [JBoss JIRA] Commented: (ISPN-1275) Virtual nodes producing same hash id for two different cluster nodes

[ https://issues.jboss.org/browse/ISPN-1275?page=com.atlassian.jira.plugin.... ] Galder Zamarreño commented on ISPN-1275: ---------------------------------------- Dan, I don't think that's the problem. From what I understand, let's say that NodeA has a set of ids. Then when Rehasher comes around, AbstractWheelConsistentHash.setCaches does not take the old hash ids into account. It simply recalculates the whole thing without taking previous hash id mappings into account, potentially leading to movements of hash ids. One way I could see this being solved is by keeping a copy of the positions and in getNormalizedHash() checking that we're not overriding an old position. That is of couse, assuming my assumption is still valid. Otherwise, there is an alternative that involves Hot Rod going through all nodes in a given view, and calling getHashIds() for them whenever a new node starts up. This performance operation is dependant on the number of nodes in the cluster though which is not nice. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira