[
https://issues.apache.org/jira/browse/HBASE-9155?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13765073#comment-13765073
]
Sergey Shelukhin commented on HBASE-9155:
-----------------------------------------
Ok, I started trying to merge my patch into it... it looks like I will just
discard my updaterthread/deleterthread inside MultiWriter, because they are
very similar to what this patch does. I think it would make more sense to have
threads inside writer as opposed to separate class for that but it's probably
too late/unimportant to change now.
One question I have is about correctness of verification in case of
deletes/increments/appends.
Original test used to fail when multiple CFs were present (e.g.
IntegrationTestLazyCf...) because read could start verifying and read partial
view of the row, not all CFs were updated; isMultiPut fixed this.
How is the reader here prevented from reading partial view in case of updates -
increments/etc. are sent separately, and MUTATE_INFO is updated later, so
reader could read intermediate version.
> Enhance LoadTestTool to support updates
> ---------------------------------------
>
> Key: HBASE-9155
> URL: https://issues.apache.org/jira/browse/HBASE-9155
> Project: HBase
> Issue Type: Test
> Reporter: Jimmy Xiang
> Assignee: Jimmy Xiang
> Fix For: 0.98.0, 0.95.2
>
> Attachments: trunk-9155.patch
>
>
> We can enhance LoadTestTool to support updates like
> increment/append/checkAndPut/checkAndDelete to make sure these operations
> work as expected during integration testing.
> If chaosmonkey is enabled, increment/append could be invoked multiple times.
> During verification, we need to keep this in mind before they are guaranteed
> to be called only once in the sever side.
--
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
