[
https://issues.apache.org/jira/browse/BEAM-1531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16067336#comment-16067336
]
Ismaël Mejía commented on BEAM-1531:
------------------------------------
Oh wow, that's an interesting approach, I like the fact that it can be useful
to have something like the fake service implementation (as BigtableIO does),
however I am not sure if we win a lot with this, I mean I see that this for
sure will improve the time of the test execution but from the test accuracy
point of view my confidence with the embedded HBase server version is higher
(and also we cover the other cases split/filters, etc), but well I am still in
doubt.
I saw that with your changes the speed of testReadingSplitAtFraction really
goes down, however I still don't understand why the other implementation takes
so long, do you have any idea on this ?
> Support dynamic work rebalancing for HBaseIO
> --------------------------------------------
>
> Key: BEAM-1531
> URL: https://issues.apache.org/jira/browse/BEAM-1531
> Project: Beam
> Issue Type: Improvement
> Components: sdk-java-extensions
> Reporter: Ismaël Mejía
> Assignee: Ismaël Mejía
> Priority: Minor
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
