[jira] [Commented] (SPARK-54640) Replace select.select with select.poll in worker

["Wojciech Szlachta (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-54640?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18043877#comment-18043877 ] 

Wojciech Szlachta commented on SPARK-54640:
-------------------------------------------

Duplicate of SPARK-51966

I think this only partially fixes the issue as similar can occur in `python/pyspark/accumulators.py`, where polling function uses `select.select()` and can fail for file descriptor numbers >=1024.

This was addressed in: https://github.com/apache/spark/pull/53306

> Replace select.select with select.poll in worker
> ------------------------------------------------
>
>                 Key: SPARK-54640
>                 URL: https://issues.apache.org/jira/browse/SPARK-54640
>             Project: Spark
>          Issue Type: Improvement
>          Components: PySpark
>    Affects Versions: 4.2.0
>            Reporter: Tian Gao
>            Assignee: Tian Gao
>            Priority: Minor
>              Labels: pull-request-available
>             Fix For: 4.2.0
>
>
> We use select.select in worker.py to communicate with daemon. However, select.select has a limit that it can't deal with fd > 1024. This might be an issue on a heavy load system. We should use the more modern select.poll to do this.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org