flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chesnay Schepler (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-10214) why is job applying as many TMs as default parallelism when starting, each parallelism is 1
Date Tue, 28 Aug 2018 07:59:00 GMT

    [ https://issues.apache.org/jira/browse/FLINK-10214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16594673#comment-16594673
] 

Chesnay Schepler commented on FLINK-10214:
------------------------------------------

As outlined in FLINK-9455 this behavior was introduced in 1.5, and is planned to be fixed
for 1.7. If this is a serious issue for you, please consider switching the cluster to legacy
mode. https://ci.apache.org/projects/flink/flink-docs-master/ops/config.html#mode

> why is job applying as many TMs as default parallelism when starting, each parallelism
is 1
> -------------------------------------------------------------------------------------------
>
>                 Key: FLINK-10214
>                 URL: https://issues.apache.org/jira/browse/FLINK-10214
>             Project: Flink
>          Issue Type: Task
>          Components: Cluster Management, TaskManager, YARN
>    Affects Versions: 1.5.0
>            Reporter: lzh9
>            Priority: Major
>
> If I set TM number=6, slot num each TM=8, memory each TM=4G, the job will apply 48 TMs,
and use only one slot each, total memory usage is 192G. A few minutes later, it will release
40TMs, everything will be normal. 
> Is there any reason about it? 



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message