[
https://issues.apache.org/jira/browse/SAMZA-2160?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hai Lu updated SAMZA-2160:
--------------------------
Fix Version/s: 1.3
> Ability to configure application type for a YARN deploy
> -------------------------------------------------------
>
> Key: SAMZA-2160
> URL: https://issues.apache.org/jira/browse/SAMZA-2160
> Project: Samza
> Issue Type: Improvement
> Reporter: Abhishek Shivanna
> Assignee: Abhishek Shivanna
> Priority: Trivial
> Fix For: 1.3
>
> Attachments: image-2019-04-08-14-47-13-740.png
>
> Time Spent: 2h 50m
> Remaining Estimate: 0h
>
> With the introduction of new API types (ex: HighLevel, SQL, Beam) it would be
> useful to distinguish/tag job types when they are deployed on YARN.
> This metadata can be used by tooling to distinguish between different types
> of jobs. Setting the ApplicationType enables us to tag deployments without
> needing external state to store this.
> Currently all jobs deployed on YARN are tagged as "Samza"
> eg:
> !image-2019-04-08-14-47-13-740.png!
> With this change the idea is to have the ability to tag deployments via a
> config (if needed)
> eg:
> {noformat}
> yarn.application.type=SamzaSQL{noformat}
> and calls to the YARN rest endpoints to filter by type would be similar to:
> {noformat}
> curl -L "http://<rm http
> address:port>/ws/v1/cluster/apps?applicationTypes=SamzaSQL"{noformat}
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
