You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Mridul Verma (JIRA)" <ji...@apache.org> on 2019/07/03 11:10:00 UTC

[jira] [Created] (FLINK-13080) Implement Sparrow Scheduler for Flink

Mridul Verma created FLINK-13080:
------------------------------------

             Summary: Implement Sparrow Scheduler for Flink
                 Key: FLINK-13080
                 URL: https://issues.apache.org/jira/browse/FLINK-13080
             Project: Flink
          Issue Type: Bug
          Components: flink-contrib
            Reporter: Mridul Verma


Sparrow Scheduler have various benefits as compared to current schedulers
 * With sparrow, we will have no common shared state for the cluster tasks. This essentially means that scheduler will be scalable in nature. Now any node can schedule the task.
 * It is meant for clusters with large number of nodes. Theoretically most of the times, it seems to perform better than the current schedulers.

[https://github.com/radlab/sparrow]



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