[
https://issues.apache.org/jira/browse/SPARK-24918?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16571748#comment-16571748
]
Luca Canali commented on SPARK-24918:
-------------------------------------
I have a use case where I would like to sample stack traces of the Spark
executors across the cluster and later aggregate the data into a Flame Graph. I
may want to do data collection only for a short duration (due to the overhead)
and possibly be able to start and stop data collection at will from the driver.
Similar use cases would be to deploy "probes" using tools for dynamic tracing
to measure specific details of the workload.
I think the executor plugin would be useful for this. In additional it would
be nice to have a mechanism to send and receive commands/data between the Spark
driver and the plugin process.
Would this proposal make sense in the context of this SPIP or would it add too
much complexity to the original proposal?
> Executor Plugin API
> -------------------
>
> Key: SPARK-24918
> URL: https://issues.apache.org/jira/browse/SPARK-24918
> Project: Spark
> Issue Type: New Feature
> Components: Spark Core
> Affects Versions: 2.4.0
> Reporter: Imran Rashid
> Priority: Major
> Labels: SPIP, memory-analysis
>
> It would be nice if we could specify an arbitrary class to run within each
> executor for debugging and instrumentation. Its hard to do this currently
> because:
> a) you have no idea when executors will come and go with DynamicAllocation,
> so don't have a chance to run custom code before the first task
> b) even with static allocation, you'd have to change the code of your spark
> app itself to run a special task to "install" the plugin, which is often
> tough in production cases when those maintaining regularly running
> applications might not even know how to make changes to the application.
> For example, https://github.com/squito/spark-memory could be used in a
> debugging context to understand memory use, just by re-running an application
> with extra command line arguments (as opposed to rebuilding spark).
> I think one tricky part here is just deciding the api, and how its versioned.
> Does it just get created when the executor starts, and thats it? Or does it
> get more specific events, like task start, task end, etc? Would we ever add
> more events? It should definitely be a {{DeveloperApi}}, so breaking
> compatibility would be allowed ... but still should be avoided. We could
> create a base class that has no-op implementations, or explicitly version
> everything.
> Note that this is not needed in the driver as we already have SparkListeners
> (even if you don't care about the SparkListenerEvents and just want to
> inspect objects in the JVM, its still good enough).
--
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
