Zhijie Shen commented on YARN-3047:

Yeah, I did {{patch -p0}}, but applying failed at {{yarn.cmd}}. [~gtCarrera9], 
does this patch work on you box?

bq. My assumption here is that we only care about per-application partial 
order, but not a total order on the timeline server. 

IMHO, we may care about total order. To be more accurate, it may not be 
"total", but among the applications of the same flow. I think eventually we 
should support reading from collector to get up-to-date/realtime timeline data.

bq. If later we deem that the view is too stale and we need to get the more 
recent data from the timeline collector, we could add that functionality in a 
separate effort.

Exactly, I agree with this prioritization. Let's start from reading the 
backend, and we will get a better sense how stale the timeline data will be.

> [Data Serving] Set up ATS reader with basic request serving structure and 
> lifecycle
> -----------------------------------------------------------------------------------
>                 Key: YARN-3047
>                 URL: https://issues.apache.org/jira/browse/YARN-3047
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: timelineserver
>            Reporter: Sangjin Lee
>            Assignee: Varun Saxena
>         Attachments: Timeline_Reader(draft).pdf, YARN-3047.001.patch, 
> YARN-3047.003.patch, YARN-3047.005.patch, YARN-3047.02.patch, 
> YARN-3047.04.patch
> Per design in YARN-2938, set up the ATS reader as a service and implement the 
> basic structure as a service. It includes lifecycle management, request 
> serving, and so on.

This message was sent by Atlassian JIRA

Reply via email to