You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Benjamin Lerer (Jira)" <ji...@apache.org> on 2020/11/02 13:14:00 UTC

[jira] [Commented] (CASSANDRA-16186) SEPExecutor does not release blocked threads as it should

    [ https://issues.apache.org/jira/browse/CASSANDRA-16186?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17224656#comment-17224656 ] 

Benjamin Lerer commented on CASSANDRA-16186:
--------------------------------------------

|| Branch || CI runs ||
| [cassandra-3.0|https://github.com/apache/cassandra/pull/799]| [J8|https://app.circleci.com/pipelines/github/blerer/cassandra?branch=CASSANDRA-16186-3.0]|
|[cassandra-3.11|https://github.com/apache/cassandra/pull/800]| [J8|https://app.circleci.com/pipelines/github/blerer/cassandra?branch=CASSANDRA-16186-3.11]|
|[trunk|https://github.com/apache/cassandra/pull/775]| [J8|https://app.circleci.com/pipelines/github/blerer/cassandra?branch=CASSANDRA-16186-3.11]|[J8|https://app.circleci.com/pipelines/github/blerer/cassandra/52/workflows/98ad6dde-bbf9-46c5-8fe6-9187e28edcc5], [J11|https://app.circleci.com/pipelines/github/blerer/cassandra/52/workflows/ce514222-6501-4b2f-8481-f0f00dc7cc31]|

As the maxTasksQueued was always set to INTEGER.MAX_VALUE, the code was in practice never using the blocking logic. By consequence, removing the code was a better option than trying to fix it.

The trunk PR add some tests for the ThreadPoolMetrics.

> SEPExecutor does not release blocked threads as it should
> ---------------------------------------------------------
>
>                 Key: CASSANDRA-16186
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-16186
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Test/unit
>            Reporter: Benjamin Lerer
>            Assignee: Benjamin Lerer
>            Priority: Normal
>             Fix For: 4.0-beta
>
>
> While adding some tests for the {{ThreadPoolMetrics}}, I discovered that the {{SEPExecutor}} does not release the blocked threads as it should.
> If the number of tasks added to a SEPExecutor exceed the max queue size. The threads adding those task will be block until enough space become available for all the blocked tasks. At this point all the blocked threads will released at once. 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org