[jira] [Updated] (FLINK-22305) Improve log messages of sort-merge blocking shuffle
[ https://issues.apache.org/jira/browse/FLINK-22305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Yingjie Cao updated FLINK-22305: Summary: Improve log messages of sort-merge blocking shuffle (was: Improve log message of sort-merge blocking shuffle) > Improve log messages of sort-merge blocking shuffle > --- > > Key: FLINK-22305 > URL: https://issues.apache.org/jira/browse/FLINK-22305 > Project: Flink > Issue Type: Sub-task > Components: Runtime / Network >Reporter: Yingjie Cao >Priority: Major > Fix For: 1.13.0 > > > Currently, the default value of taskmanager.network.sort-shuffle.min-buffers > is 64, which is pretty small. As suggested, we'd like to increase the default > value of taskmanager.network.sort-shuffle.min-buffers. By increasing the > default taskmanager.network.sort-shuffle.min-buffers, the corner case of very > small in-memory sort-buffer and write-buffer can be avoid, which is better > for performance. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Updated] (FLINK-22305) Improve log messages of sort-merge blocking shuffle
[ https://issues.apache.org/jira/browse/FLINK-22305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ASF GitHub Bot updated FLINK-22305: --- Labels: pull-request-available (was: ) > Improve log messages of sort-merge blocking shuffle > --- > > Key: FLINK-22305 > URL: https://issues.apache.org/jira/browse/FLINK-22305 > Project: Flink > Issue Type: Sub-task > Components: Runtime / Network >Reporter: Yingjie Cao >Priority: Major > Labels: pull-request-available > Fix For: 1.13.0 > > > Currently, the default value of taskmanager.network.sort-shuffle.min-buffers > is 64, which is pretty small. As suggested, we'd like to increase the default > value of taskmanager.network.sort-shuffle.min-buffers. By increasing the > default taskmanager.network.sort-shuffle.min-buffers, the corner case of very > small in-memory sort-buffer and write-buffer can be avoid, which is better > for performance. -- This message was sent by Atlassian Jira (v8.3.4#803005)