[
https://issues.apache.org/jira/browse/BEAM-4783?focusedWorklogId=144297&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-144297
]
ASF GitHub Bot logged work on BEAM-4783:
----------------------------------------
Author: ASF GitHub Bot
Created on: 14/Sep/18 14:26
Start Date: 14/Sep/18 14:26
Worklog Time Spent: 10m
Work Description: chamikaramj commented on issue #6181: [BEAM-4783] Add
bundleSize for splitting BoundedSources.
URL: https://github.com/apache/beam/pull/6181#issuecomment-421375360
Sorry missed this PR.
Had a quick look.
I think proper solution is to introduce dynamic work rebalancing [1] to
SparkRunner at some point. This way large bundles can be broken up into smaller
bundles if there are more workers to process work. I agree with Ismaël that
proposed solution go against the Beam's no-knobs philosophy but I understand
why it might be needed till SparkRunner has support for dynamic work
rebalancing. I'd suggest performing some experimentation to make sure that the
new option helps before introducing it.
[1]
https://cloud.google.com/blog/products/gcp/no-shard-left-behind-dynamic-work-rebalancing-in-google-cloud-dataflow
----------------------------------------------------------------
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:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 144297)
Time Spent: 2h (was: 1h 50m)
> Spark SourceRDD Not Designed With Dynamic Allocation In Mind
> ------------------------------------------------------------
>
> Key: BEAM-4783
> URL: https://issues.apache.org/jira/browse/BEAM-4783
> Project: Beam
> Issue Type: Improvement
> Components: runner-spark
> Affects Versions: 2.5.0
> Reporter: Kyle Winkelman
> Assignee: Jean-Baptiste Onofré
> Priority: Major
> Labels: newbie
> Time Spent: 2h
> Remaining Estimate: 0h
>
> When the spark-runner is used along with the configuration
> spark.dynamicAllocation.enabled=true the SourceRDD does not detect this. It
> then falls back to the value calculated in this description:
> // when running on YARN/SparkDeploy it's the result of max(totalCores,
> 2).
> // when running on Mesos it's 8.
> // when running local it's the total number of cores (local = 1,
> local[N] = N,
> // local[*] = estimation of the machine's cores).
> // ** the configuration "spark.default.parallelism" takes precedence
> over all of the above **
> So in most cases this default is quite small. This is an issue when using a
> very large input file as it will only get split in half.
> I believe that when Dynamic Allocation is enable the SourceRDD should use the
> DEFAULT_BUNDLE_SIZE and possibly expose a SparkPipelineOptions that allows
> you to change this DEFAULT_BUNDLE_SIZE.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
