hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-2195) Support cyclic replication
Date Wed, 31 Aug 2011 03:39:10 GMT

    [ https://issues.apache.org/jira/browse/HBASE-2195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13094282#comment-13094282

Lars Hofhansl commented on HBASE-2195:

There is a good chance that we will need this at Salesforce.com.

After looking at the code, the only missing piece is retaining the originating cluster information
at the ReplicationSink. (Unless, of course, I am missing something).

I have some test code hacked up, which allows setting the cluster id as a Put/Delete attribute
and then in HRegion writes the ClusterId into the WAL.
The replication source checks the HLog key and does not replicate when the cluster id equals
the peer's cluster id, in addition if the cluster id was set in the WAL it is retained.

The basic problem is that this requires core changes (HRegion), and replication so far is
nicely out of core.

> Support cyclic replication
> --------------------------
>                 Key: HBASE-2195
>                 URL: https://issues.apache.org/jira/browse/HBASE-2195
>             Project: HBase
>          Issue Type: Sub-task
>          Components: replication
>            Reporter: Jean-Daniel Cryans
> We need to support cyclic replication by using the cluster id of each HlogKey and stop
replicating when it goes back to the original cluster.

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


View raw message