[
https://issues.apache.org/jira/browse/HBASE-18183?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16178657#comment-16178657
]
Anoop Sam John commented on HBASE-18183:
----------------------------------------
[~uagashe]
We can not deprecate RowProcessor.. Pls see all usage eg:s of this. So this
RowProcessor as such is required with its pre/post/process flows.
Region#mutateRowsWithLocks and Region#processRowsWithLocks are implemented via
RowProcessor way. And making use in MultiRowMutationEndpoint. This is kind
of sample RowProcessor for this kind of multi row atomic mutate.
So even if we have to avoid all the methods in RowProcessor specific for diff
CP hooks call, we have to first break the MultiRowMutationEndpoint and
Region#mutateRowsWithLocks making use of this method.
I agree the way we have extra CP specific methods in RowProcessor and usage of
these in Region#processRowsWithLocks making us confuse a lot and easily missing
some changes so that we are facing issues like inconsistency for this API with
doMiniBatchMutate(). But the fix may be not that simple as removing
RowProcessor! :-(
> Region interface cleanup for CP expose
> --------------------------------------
>
> Key: HBASE-18183
> URL: https://issues.apache.org/jira/browse/HBASE-18183
> Project: HBase
> Issue Type: Sub-task
> Components: Coprocessors
> Reporter: Anoop Sam John
> Assignee: Anoop Sam John
> Fix For: 2.0.0-alpha-4
>
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
