Wangda Tan commented on YARN-3849:

Thanks [~sunilg],
Some comments:

Some comments:
1) It seems we don't need useDominantResourceCalculator/rcDefault/rcDominant in 
TestP..Policy, pass a boolean parameter to buildPolicy should be enough, you 
can also overload a buildPolicy to avoid too much changes.

2) testPreemptionWithVCoreResource seems not correct, root.used != A.used + 

3) TestP..PolicyFroNodePartitions:
One comments is wrong:
        + "(1,1:2,n1,x,20,false);" + // 80 * x in n1
        "b\t" // app4 in b
        + "(1,1:2,n2,,80,false)"; // 20 default in n2
It should be 20 * x and 80 default

4) It seems TestP..PolicyFroNodePartitions setting for DRC is missing, could 
you check?

> Too much of preemption activity causing continuos killing of containers 
> across queues
> -------------------------------------------------------------------------------------
>                 Key: YARN-3849
>                 URL: https://issues.apache.org/jira/browse/YARN-3849
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: capacityscheduler
>    Affects Versions: 2.7.0
>            Reporter: Sunil G
>            Assignee: Sunil G
>            Priority: Critical
>         Attachments: 0001-YARN-3849.patch, 0002-YARN-3849.patch
> Two queues are used. Each queue has given a capacity of 0.5. Dominant 
> Resource policy is used.
> 1. An app is submitted in QueueA which is consuming full cluster capacity
> 2. After submitting an app in QueueB, there are some demand  and invoking 
> preemption in QueueA
> 3. Instead of killing the excess of 0.5 guaranteed capacity, we observed that 
> all containers other than AM is getting killed in QueueA
> 4. Now the app in QueueB is trying to take over cluster with the current free 
> space. But there are some updated demand from the app in QueueA which lost 
> its containers earlier, and preemption is kicked in QueueB now.
> Scenario in step 3 and 4 continuously happening in loop. Thus none of the 
> apps are completing.

This message was sent by Atlassian JIRA

Reply via email to