[
https://issues.apache.org/jira/browse/HBASE-18906?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16191702#comment-16191702
]
Sergey Soldatov commented on HBASE-18906:
-----------------------------------------
[~elserj] Sure. Why we need this hack? That's simple. For some cases such as
index creation or upserting data basing on the existing data (upsert select
queries) we do scan and batch writes in the coprocessor hook. That lead to very
high write load and region server just unable to handle all those writes and
fail with region too busy/not serving exceptions. Calling flush directly is not
the case and makes everything even worst. That terrible code is supposed to
throttle when memstore reached some watermark, so RS will be able to process
flushes in regular way. Requesting flush async and ability to wait for it's
completion would definitely help us a lot.
> Investigate Phoenix usages of Region#waitForXXX APIs
> ----------------------------------------------------
>
> Key: HBASE-18906
> URL: https://issues.apache.org/jira/browse/HBASE-18906
> Project: HBase
> Issue Type: Sub-task
> Components: Coprocessors
> Reporter: Anoop Sam John
> Assignee: Anoop Sam John
> Fix For: 2.0.0-alpha-4
>
>
> While reviewing HBASE-18183, Andy pointed out that Phoenix uses
> waitForFlushesAndCompactions and/or waitForFlushes for diff reasons. This
> issue is to see why they need them and whether alternate ways are possible.
> This seems to be too much internal stuff and a normal CP hook calling these
> would be dangerous.
> If there are alternate ways for Phoenix not to use this and not landing in
> issues (As said by Andy) we should suggest/fix for them.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
