Karthik Kambatla commented on YARN-1879:

Thanks for looking at it closely, Jian and Xuan. I have missed some of these 
points. Spent a little more time thinking through.

bq. I think we are mixing two issues in this jira
When we mark an API Idempotent or AtMostOnce, the retry-policies will end up 
re-invoking the API on the other RM in case of a failover. This is okay only if 
the RM handles these duplicate requests. Further, my understanding is that the 
behavior of "Idempotent" APIs should be the same on each invocation; i.e., the 
client should receive the exact same response too. 

If we handle duplicate requests but return a different response to the client 
on duplicate calls, we can mark it AtMostOnce. If we return the same response, 
we can go ahead and mark it Idempotent. Needless to say, the RM should 
definitely duplicate requests gracefully. Does that sound reasonable? 

> Mark Idempotent/AtMostOnce annotations to ApplicationMasterProtocol for RM 
> fail over
> ------------------------------------------------------------------------------------
>                 Key: YARN-1879
>                 URL: https://issues.apache.org/jira/browse/YARN-1879
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: resourcemanager
>            Reporter: Jian He
>            Assignee: Tsuyoshi OZAWA
>            Priority: Critical
>         Attachments: YARN-1879.1.patch, YARN-1879.1.patch, 
> YARN-1879.11.patch, YARN-1879.12.patch, YARN-1879.13.patch, 
> YARN-1879.14.patch, YARN-1879.15.patch, YARN-1879.16.patch, 
> YARN-1879.17.patch, YARN-1879.18.patch, YARN-1879.19.patch, 
> YARN-1879.2-wip.patch, YARN-1879.2.patch, YARN-1879.20.patch, 
> YARN-1879.21.patch, YARN-1879.22.patch, YARN-1879.3.patch, YARN-1879.4.patch, 
> YARN-1879.5.patch, YARN-1879.6.patch, YARN-1879.7.patch, YARN-1879.8.patch, 
> YARN-1879.9.patch

This message was sent by Atlassian JIRA

Reply via email to