[
https://issues.apache.org/jira/browse/HIVE-23166?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Panagiotis Garefalakis updated HIVE-23166:
------------------------------------------
Status: Patch Available (was: In Progress)
> Guard VGB from flushing too often
> ---------------------------------
>
> Key: HIVE-23166
> URL: https://issues.apache.org/jira/browse/HIVE-23166
> Project: Hive
> Issue Type: Improvement
> Components: llap
> Affects Versions: 4.0.0
> Reporter: Panagiotis Garefalakis
> Assignee: Panagiotis Garefalakis
> Priority: Major
> Attachments: HIVE-23166.01.patch
>
>
> The existing flush logic in our VectorGroupByOperator is completely static.
> It depends on the: number of HtEntries
> (*hive.vectorized.groupby.maxentries*) and the MAX memory threshold (by
> default 90% of available memory)
>
> Assuming that we are not memory constrained the periodicity of flushing is
> currently dictated by the static number of entries (1M by default) which can
> be also misconfigured to a very low value.
> I am proposing along with maxHtEntries, to also take into account current
> memory usage, to avoid flushing too ofter as it can hurt op throughput for
> particular workloads.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
