[ 
https://issues.apache.org/jira/browse/YARN-2628?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14155584#comment-14155584
 ] 

Jian He commented on YARN-2628:
-------------------------------

looks good, one minor comment in the test case:
- the following assertion depends on timing, as the allocation happens 
asynchronously, it might fail. could you use a loop to check if the container 
is allocated, otherwise timeout.
{code}
    Thread.sleep(1000);
    allocResponse = am1.schedule();
    Assert.assertEquals(1, allocResponse.getAllocatedContainers().size());
{code}

> Capacity scheduler with DominantResourceCalculator carries out reservation 
> even though slots are free
> -----------------------------------------------------------------------------------------------------
>
>                 Key: YARN-2628
>                 URL: https://issues.apache.org/jira/browse/YARN-2628
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: capacityscheduler
>    Affects Versions: 2.5.1
>            Reporter: Varun Vasudev
>            Assignee: Varun Vasudev
>         Attachments: apache-yarn-2628.0.patch
>
>
> We've noticed that if you run the CapacityScheduler with the 
> DominantResourceCalculator, sometimes apps will end up with containers in a 
> reserved state even though free slots are available.
> The root cause seems to be this piece of code from CapacityScheduler.java -
> {noformat}
>     // Try to schedule more if there are no reservations to fulfill
>     if (node.getReservedContainer() == null) {
>       if (Resources.greaterThanOrEqual(calculator, getClusterResource(),
>           node.getAvailableResource(), minimumAllocation)) {
>         if (LOG.isDebugEnabled()) {
>           LOG.debug("Trying to schedule on node: " + node.getNodeName() +
>               ", available: " + node.getAvailableResource());
>         }
>         root.assignContainers(clusterResource, node, false);
>       }
>     } else {
>       LOG.info("Skipping scheduling since node " + node.getNodeID() + 
>           " is reserved by application " + 
>           
> node.getReservedContainer().getContainerId().getApplicationAttemptId()
>           );
>     }
> {noformat}
> The code is meant to check if a node has any slots available for containers . 
> Since it uses the greaterThanOrEqual function, we end up in situation where 
> greaterThanOrEqual returns true, even though we may not have enough CPU or 
> memory to actually run the container.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to