You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Dongjoon Hyun (Jira)" <ji...@apache.org> on 2024/01/25 16:41:00 UTC
[jira] [Commented] (SPARK-46861) Avoid Deadlock in DAGScheduler
[ https://issues.apache.org/jira/browse/SPARK-46861?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17810972#comment-17810972 ]
Dongjoon Hyun commented on SPARK-46861:
---------------------------------------
I added a label, `releasenotes`, because this is an important bug fix.
> Avoid Deadlock in DAGScheduler
> ------------------------------
>
> Key: SPARK-46861
> URL: https://issues.apache.org/jira/browse/SPARK-46861
> Project: Spark
> Issue Type: Bug
> Components: Scheduler
> Affects Versions: 4.0.0, 3.5.1, 3.4.3
> Reporter: Fredrik Klauß
> Assignee: Fredrik Klauß
> Priority: Major
> Labels: pull-request-available, releasenotes
> Fix For: 4.0.0, 3.5.1, 3.4.3
>
>
> * The DAGScheduler could currently run into a deadlock with another thread if both access the partitions of the same RDD at the same time.
> * To make progress in getCacheLocs, we require both exclusive access to the RDD partitions and the location cache. We first lock on the location cache, and then on the RDD.
> * When accessing partitions of an RDD, the RDD first acquires exclusive access on the partitions, and then might acquire exclusive access on the location cache.
> * If thread 1 is able to acquire access on the RDD, while thread 2 holds the access to the location cache, we can run into a deadlock situation.
--
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