[ 
https://issues.apache.org/jira/browse/YARN-1408?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14019783#comment-14019783
 ] 

Sunil G commented on YARN-1408:
-------------------------------

Hi [~leftnoteasy]
bq.I meant we need keep a complete Resource Request, which includes Resource 
Request itself, rack/ANY level Resource Request, in case of the entire 
ResourceRequest removed when count reaches zero. I think we don't have 
difference here. , right?
Yes. :)

bq.pass a reference of SchedulerApplicationAttempt to RMContainer
My doubt is in this context. I felt like we may not need a 
SchedulerApplicationAttempt in RMContainer itself. Rather I can do like keeping 
required information in RMContainer. So it will be more sepcific to the 
handling.

bq.IMHO, it's better to put recover ResourceRequest logic in 
RMContainerImpl.FinishedTransition(), we can check if original state is 
Allocated and event is KILL. The benefit of this choice is we don't need 
separately modify FairScheduler and CapacityScheduler. Make sense?
KILL event can happen not only from Preemption. I feel we need to deal issue 
with repect to preemption only, correct?. Other KILL event in RMContainer may 
be for a reserved container KILL etc.
So we may need to do only in PreemptableResourceScheduler#killContainer. What 
do you feel?

bq.Yes, but we cannot reuse this API without modifying its definition. The 
AppSchedulingInfo#updateResourceRequests will replace original ResourceRequests 
to new ResourceRequests, and our requirement is only increase the original 
ResourceRequest.
I agree with the concern of modifying definition, but to an extent I think it 
will be better if we can reuse with a recovery mode.

bq.And please note that, we need update QueueMetrics as well.
+1 Yes. I agree. We have to do this also.

Thank you. 
Sunil

> Preemption caused Invalid State Event: ACQUIRED at KILLED and caused a task 
> timeout for 30mins
> ----------------------------------------------------------------------------------------------
>
>                 Key: YARN-1408
>                 URL: https://issues.apache.org/jira/browse/YARN-1408
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: resourcemanager
>    Affects Versions: 2.2.0
>            Reporter: Sunil G
>            Assignee: Sunil G
>         Attachments: Yarn-1408.1.patch, Yarn-1408.2.patch, Yarn-1408.3.patch, 
> Yarn-1408.4.patch, Yarn-1408.patch
>
>
> Capacity preemption is enabled as follows.
>  *  yarn.resourcemanager.scheduler.monitor.enable= true ,
>  *  
> yarn.resourcemanager.scheduler.monitor.policies=org.apache.hadoop.yarn.server.resourcemanager.monitor.capacity.ProportionalCapacityPreemptionPolicy
> Queue = a,b
> Capacity of Queue A = 80%
> Capacity of Queue B = 20%
> Step 1: Assign a big jobA on queue a which uses full cluster capacity
> Step 2: Submitted a jobB to queue b  which would use less than 20% of cluster 
> capacity
> JobA task which uses queue b capcity is been preempted and killed.
> This caused below problem:
> 1. New Container has got allocated for jobA in Queue A as per node update 
> from an NM.
> 2. This container has been preempted immediately as per preemption.
> Here ACQUIRED at KILLED Invalid State exception came when the next AM 
> heartbeat reached RM.
> ERROR 
> org.apache.hadoop.yarn.server.resourcemanager.rmcontainer.RMContainerImpl: 
> Can't handle this event at current state
> org.apache.hadoop.yarn.state.InvalidStateTransitonException: Invalid event: 
> ACQUIRED at KILLED
> This also caused the Task to go for a timeout for 30minutes as this Container 
> was already killed by preemption.
> attempt_1380289782418_0003_m_000000_0 Timed out after 1800 secs



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Reply via email to