You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flink.apache.org by "bo zhao (Jira)" <ji...@apache.org> on 2022/08/05 07:42:00 UTC

[jira] [Created] (FLINK-28833) [5] Support customized scheduler options as a hashmap for each scheduler.

bo zhao created FLINK-28833:
-------------------------------

             Summary: [5] Support customized scheduler options as a hashmap for each scheduler.
                 Key: FLINK-28833
                 URL: https://issues.apache.org/jira/browse/FLINK-28833
             Project: Flink
          Issue Type: Sub-task
          Components: Deployment / Kubernetes
            Reporter: bo zhao


-D{*}kubernetes.scheduler.<scheduler-name>.config=priorityclass:hight-priority{*}{*},X:Y…{*}

The option value is a map, and will be loaded by the specified customized scheduler.

Additional K8S customized configuration proposal, such as PodGroup in Volcano:
|*Key*|*Category*|*Related Config Options*|*Description*|
|config|Defined by user|kubernetes.scheduler.<scheduler-name>.config|The customized Pod Group definition for JobManager and TaskManager. Such as:
mincpu:1, minmember:2, minmemory:1000Mi, priorityclass: <CLASSNAME>, the customized scheduler will get the message it is concerned, ignoring others. |

 



--
This message was sent by Atlassian Jira
(v8.20.10#820010)