[
https://issues.apache.org/jira/browse/SPARK-30299?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17076525#comment-17076525
]
Thomas Graves commented on SPARK-30299:
---------------------------------------
Note that there are other places in the code that uses executor cores which
could also be wrong in standalone mode. for instance PythonRunner is using it
to split memory.
> Dynamic allocation with Standalone mode calculates to many executors needed
> ---------------------------------------------------------------------------
>
> Key: SPARK-30299
> URL: https://issues.apache.org/jira/browse/SPARK-30299
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 2.4.4, 3.0.0
> Reporter: Thomas Graves
> Priority: Major
>
> While I was doing some changes in the executor allocation manager, I realized
> there is a bug with dynamic allocation in standalone mode.
> The issue is that if you run standalone mode with the default settings where
> the executor gets all the cores of the worker, spark core (allocation
> manager) doesn't know the number of cores per executor to be able to
> calculate how many tasks can fit on an executor.
> It therefore defaults to the use the default EXECUTOR_CORES which is 1 and
> thus could calculate it needs way more containers there are actually does.
> For instance, I have a worker with 12 cores. That means by default when I
> start an executor on it, it gets 12 cores and can fit 12 tasks. The
> allocation manager would use the default of 1 core per executor and say it
> needs 12 executors when it only needs 1.
> The fix for this isn't trivial since it would need to know how many cores
> each one has and I assume it would also need to handle heterogenous nodes.
> I could start workers on nodes with different numbers of cores - one with 24
> cores and one with 16 cores. How do we estimate the number of executors in
> this case. We could just choose the min of existing ones or something like
> that as an estimate and it would be closer, unless of course the next
> executor you got didn't actually have that.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
