[
https://issues.apache.org/jira/browse/HBASE-10576?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13978294#comment-13978294
]
James Taylor commented on HBASE-10576:
--------------------------------------
Let me add a (4) too, related to my question above: How would the client know
that a scan was serviced by the shadow region instead of the regular data
region? This is important since the client might otherwise be forced to reorder
all the rows it gets back versus doing a merge sort among a set of parallel
scans. Also related (and perhaps (5)), could separate shadow regions be used
per index (i.e. would that scale fine)? If not, then the row key of the shadow
region would need to be prefixed in some way. In this case, how would the
client and server-side code deal with this?
I think this idea sound very promising, but I think the APIs would likely need
to be tweaked so that the client has visibility into the usage of these shadow
regions.
> Custom load balancer to co-locate the regions of two tables which are having
> same split keys
> --------------------------------------------------------------------------------------------
>
> Key: HBASE-10576
> URL: https://issues.apache.org/jira/browse/HBASE-10576
> Project: HBase
> Issue Type: Sub-task
> Components: Balancer
> Reporter: rajeshbabu
> Assignee: rajeshbabu
> Attachments: HBASE-10536_v2.patch, HBASE-10576.patch
>
>
> To support local indexing both user table and index table should have same
> split keys. This issue to provide custom balancer to colocate the regions of
> two tables which are having same split keys.
> This helps in Phoenix as well.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
