[
https://issues.apache.org/jira/browse/FLINK-22037?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Zhilong Hong updated FLINK-22037:
---------------------------------
Description:
We find that the {{BlockingQueue}} that used to preserve
{{TaskDeploymentDescriptors}} which are never used later. Since a
{{TaskDeploymentDescriptor}} would cost massive heap memory, it may introduce
unnecessary garbage collection and make the result of benchmarks related to
deployment unstable. Therefore, we think it's better to remove the redundant
blocking queue in {{DeployingTasksBenchmarkBase}}.
was:We find that the {{BlockingQueue}} that used to preserve
{{TaskDeploymentDescriptors}} are never used later. Since
{{TaskDeploymentDescriptors}} would cost massive heap memory, it may introduce
unnecessary garbage collection and make the result of benchmarks related to
deployment unstable. So we think it's better to remove the redundant blocking
queue in {{DeployingTasksBenchmarkBase}}.
> Remove the redundant blocking queue from DeployingTasksBenchmarkBase
> --------------------------------------------------------------------
>
> Key: FLINK-22037
> URL: https://issues.apache.org/jira/browse/FLINK-22037
> Project: Flink
> Issue Type: Bug
> Components: Benchmarks, Runtime / Coordination
> Affects Versions: 1.13.0
> Reporter: Zhilong Hong
> Priority: Major
> Fix For: 1.13.0
>
>
> We find that the {{BlockingQueue}} that used to preserve
> {{TaskDeploymentDescriptors}} which are never used later. Since a
> {{TaskDeploymentDescriptor}} would cost massive heap memory, it may introduce
> unnecessary garbage collection and make the result of benchmarks related to
> deployment unstable. Therefore, we think it's better to remove the redundant
> blocking queue in {{DeployingTasksBenchmarkBase}}.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
