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

[Manik Surtani (JIRA)]

    [ https://issues.jboss.org/browse/ISPN-1275?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12617510#comment-12617510 ] 

Manik Surtani commented on ISPN-1275:
-------------------------------------

Right, and this is because we build a new CH based on a new view and hence the 50% chance of colliding normalised hashes being bumped inconsistently.

How about this as a fix: when we build a new CH, we do so based on the *old* CH, and then we *add* the joiners to it.  This will ensure that it is the joiners that get bumped, and will be deterministic/consistent across the cluster.  Thoughts?

> ConsistentHash implementation producing same hash id for two different cluster nodes
> ------------------------------------------------------------------------------------
>
>                 Key: ISPN-1275
>                 URL: https://issues.jboss.org/browse/ISPN-1275
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Distributed Cache
>    Affects Versions: 5.0.0.CR8
>            Reporter: Galder Zamarreño
>            Assignee: Manik Surtani
>            Priority: Blocker
>              Labels: consistent_hash
>             Fix For: 5.0.0.FINAL
>
>         Attachments: hashids.log
>
>
> While working on ISPN-1273, I've discovered that when virtual nodes are enabled, ConsistentHash.getHashIds(address) can return the same hash for two different cluster nodes. Example: with 500 virtual nodes, both NodeA-23181 and NodeB-39177 produce hash id *5289*. See attached log.
> Virtual nodes make the problem more apparent, however the problem exists regardless of vnodes being used.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira