[
https://issues.apache.org/jira/browse/SPARK-24815?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16852874#comment-16852874
]
Stavros Kontopoulos commented on SPARK-24815:
---------------------------------------------
@[~Karthik Palaniappan] I can help with the design. Btw there is a refactoring
happen in here: [https://github.com/apache/spark/pull/24704]
My concern with batch mode dynamic allocation is task list may not tell the
whole story, what if the number of tasks stays the same and load changes per
task/partition eg. Kafka source? As for state I think you need to rebalance it
which translates to dynamic re-partitioning for the micro-batch mode in
structured streaming. For continuous streaming it is harder I think, but maybe
a unified approach could solve it for both batch and continuous streaming as in
Flink, https://flink.apache.org/features/2017/07/04/flink-rescalable-state.html.
> Structured Streaming should support dynamic allocation
> ------------------------------------------------------
>
> Key: SPARK-24815
> URL: https://issues.apache.org/jira/browse/SPARK-24815
> Project: Spark
> Issue Type: Improvement
> Components: Scheduler, Structured Streaming
> Affects Versions: 2.3.1
> Reporter: Karthik Palaniappan
> Priority: Minor
>
> For batch jobs, dynamic allocation is very useful for adding and removing
> containers to match the actual workload. On multi-tenant clusters, it ensures
> that a Spark job is taking no more resources than necessary. In cloud
> environments, it enables autoscaling.
> However, if you set spark.dynamicAllocation.enabled=true and run a structured
> streaming job, the batch dynamic allocation algorithm kicks in. It requests
> more executors if the task backlog is a certain size, and removes executors
> if they idle for a certain period of time.
> Quick thoughts:
> 1) Dynamic allocation should be pluggable, rather than hardcoded to a
> particular implementation in SparkContext.scala (this should be a separate
> JIRA).
> 2) We should make a structured streaming algorithm that's separate from the
> batch algorithm. Eventually, continuous processing might need its own
> algorithm.
> 3) Spark should print a warning if you run a structured streaming job when
> Core's dynamic allocation is enabled
--
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
