[
https://issues.apache.org/jira/browse/HBASE-10576?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13980629#comment-13980629
]
rajeshbabu commented on HBASE-10576:
------------------------------------
[~jeffreyz]
bq. For example, when an index region move, there is no guarantee that data
region is also moved and opened simultaneously on the same RS.
We are handling region movements properly.
1) If regions should be balanced to other region server then custom balancer
selects both data and index regions together so both will be moved
simultaneousily.
2) If user explicitly move data/index region corresponding region will be moved
through the hooks.
3) When region server down also both data and index regions will be assigned to
some other RS simultaneously.
bq. It's possible that during an index update a region move could happen and
the update survives the region move because of retires. So there are chances
index region & data region on different RSs even within one update operation.
We are handling updates in such a way that no region will be closed in the
middle of updates. So there is no chance of inconsistencies. If the region
assignments in the middle like user region successfully assigned and index
region assignment in progress, we are not allowing puts to any region, where we
can retry updates.
Still you wanted go with shadow regions?
Thanks.
> 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)
