[
https://issues.apache.org/jira/browse/SPARK-26311?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16718107#comment-16718107
]
ASF GitHub Bot commented on SPARK-26311:
----------------------------------------
HeartSaVioR commented on issue #23260: [SPARK-26311][YARN] New feature: custom
log URL for stdout/stderr
URL: https://github.com/apache/spark/pull/23260#issuecomment-446387143
> My understanding is that this allows pointing the Spark UI directly at the
history server (old JHS or new ATS) instead of hardcoding the NM URL and
relying on the NM redirecting you, since the NM may not exist later on.
Yes, exactly. That's one of issue this patch enables to deal with, and
another one would be cluster awareness. The existence of `the clusterId of RM`
represents that YARN opens the possibility of maintaining multiple YARN
clusters and provides centralized services which operates with multiple YARN
clusters.
> when perhaps if there was a way to hook this up on the Spark history
server side only, that may be more useful.
> I think someone tried that in the past but the SHS change was very
YARN-specific, which made it kinda sub-optimal.
I agree the case is rather not against running applications but finished
applications. Currently Spark just sets executor log urls in environment at
resource manager side and uses them. The usages are broad, and not sure we can
determine which resource manager the application is based on, and whether
application is running or finished in all usages. (I'm not familiar with UI
side.) So this patch tackles the easiest way to deal with.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> [YARN] New feature: custom log URL for stdout/stderr
> ----------------------------------------------------
>
> Key: SPARK-26311
> URL: https://issues.apache.org/jira/browse/SPARK-26311
> Project: Spark
> Issue Type: Improvement
> Components: YARN
> Affects Versions: 2.4.0
> Reporter: Jungtaek Lim
> Priority: Major
>
> Spark has been setting static log URLs for YARN application, which points to
> NodeManager webapp. Normally it would work for both running apps and finished
> apps, but there're also other approaches on maintaining application logs,
> like having external log service which enables to avoid application log url
> to be a deadlink when NodeManager is not accessible. (Node decommissioned,
> elastic nodes, etc.)
> Spark can provide a new configuration for custom log url on YARN mode, which
> end users can set it properly to point application log to external log
> service.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
