hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Demai Ni (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-5002) Control replication peer per column family.
Date Tue, 11 Jun 2013 00:16:23 GMT

    [ https://issues.apache.org/jira/browse/HBASE-5002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13680080#comment-13680080
] 

Demai Ni commented on HBASE-5002:
---------------------------------

The REPLICATION_PEERS will be some thing like an array for with value ['1','2','3'], and each
entry will be the peer_id (from list_peers)?  

If implemented, how about use the REPLICATION_PEERS to replace the existing REPLICATION_SCOP?
currently 
Scope only have two values: 0-no replication; 1 - global replication. But actually, any none-zero
value will be treated as 1 - global replication. 
                
> Control replication peer per column family.
> -------------------------------------------
>
>                 Key: HBASE-5002
>                 URL: https://issues.apache.org/jira/browse/HBASE-5002
>             Project: HBase
>          Issue Type: Sub-task
>          Components: Replication
>            Reporter: Lars Hofhansl
>            Priority: Minor
>
> With HBASE-2196 in place. Would be nice if we could control per CF to which peer(s) it
is replicated.
> Could have a new option REPLICATION_PEERS, which holds a , separated list of replication
peers that this CF should be replicated to. If not given replicate to all slaves.
> This list would need to written to each log entry (could reuse WALEdit.scopes for this),
so we can decide in ReplicationSource who this entry should go to.
> Let's discuss...

--
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

Mime
View raw message