Varun Saxena commented on YARN-4224:

bq. One quick thing is about TimelineReaderContext. This class appears to have 
a lot of duplicated logic as TimelineCollectorContext, and serves as a similar 
role. Do we want to reorganize them or consolidate them?
I see. For TimelineReaderContext we would need entity type and entity id in 
addition to fields in TimelineCollectorContext. Maybe we can rename 
TimelineCollectorContext to TimelineContext class. And extend this class as 
TimelineReaderContext containing the 2 extra fields.

bq. In the original design, /flows will return a list of flows with latest 
activities on the cluster. Shall we keep this endpoint (as I noticed, /flows 
endpoint is not touched in this patch), or attach it to somewhere else in our 
Didn't quite get your question.
/flows endpoint is for querying active flows with default cluster id. And 
/clusters/\{clusterid\}/flows/ endpoint is for querying flows for specific 
cluster id.
Didn't keep a UID related endpoint for flows because cluster id by itself is a 

> Support fetching entities by UID and change the REST interface to conform to 
> current REST APIs' in YARN
> -------------------------------------------------------------------------------------------------------
>                 Key: YARN-4224
>                 URL: https://issues.apache.org/jira/browse/YARN-4224
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: timelineserver
>    Affects Versions: YARN-2928
>            Reporter: Varun Saxena
>            Assignee: Varun Saxena
>              Labels: yarn-2928-1st-milestone
>         Attachments: YARN-4224-YARN-2928.01.patch, 
> YARN-4224-feature-YARN-2928.wip.02.patch, 
> YARN-4224-feature-YARN-2928.wip.03.patch

This message was sent by Atlassian JIRA

Reply via email to