[
https://issues.apache.org/jira/browse/PHOENIX-1432?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14207913#comment-14207913
]
Samarth Jain commented on PHOENIX-1432:
---------------------------------------
I was thinking it would be good to call ThreadPoolExecutor.purge() when we end
up canceling queued tasks.
https://docs.oracle.com/javase/7/docs/api/java/util/concurrent/ThreadPoolExecutor.html#purge()
This way queue capacity/space would be available much sooner than relying on
worker threads to clean up the queue. This would need
QueryServices.getExecutor() to return a ThreadPoolExecutor instead of
ExecutorService.
> Run limit query that has only leading PK column filter serially
> ---------------------------------------------------------------
>
> Key: PHOENIX-1432
> URL: https://issues.apache.org/jira/browse/PHOENIX-1432
> Project: Phoenix
> Issue Type: Bug
> Reporter: James Taylor
> Assignee: James Taylor
> Attachments: PHOENIX-1432.patch, PHOENIX-1432_v1.patch,
> PHOENIX-1432_v2.patch
>
>
> If we have a query with a LIMIT clause and either no WHERE clause or a WHERE
> clause that only filters on the leading PK columns, we should run it
> serially. Otherwise, we're spawning too many client-side threads that likely
> won't be required.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
