[
https://issues.apache.org/jira/browse/YARN-8771?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16619835#comment-16619835
]
Wangda Tan commented on YARN-8771:
----------------------------------
Nice catch! Thanks [~Tao Yang].
Patch LGTM as well. For the test, you can check
TestCapacitySchedulerWithMultiResourceTypes as examples about how to do unit
tests for multiple resource types without adding resource-types.xml.
And I think we should put this to branch-3.1 as well.
> CapacityScheduler fails to unreserve when cluster resource contains empty
> resource type
> ---------------------------------------------------------------------------------------
>
> Key: YARN-8771
> URL: https://issues.apache.org/jira/browse/YARN-8771
> Project: Hadoop YARN
> Issue Type: Bug
> Components: capacityscheduler
> Affects Versions: 3.2.0
> Reporter: Tao Yang
> Assignee: Tao Yang
> Priority: Critical
> Attachments: YARN-8771.001.patch, YARN-8771.002.patch
>
>
> We found this problem when cluster is almost but not exhausted (93% used),
> scheduler kept allocating for an app but always fail to commit, this can
> blocking requests from other apps and parts of cluster resource can't be used.
> Reproduce this problem:
> (1) use DominantResourceCalculator
> (2) cluster resource has empty resource type, for example: gpu=0
> (3) scheduler allocates container for app1 who has reserved containers and
> whose queue limit or user limit reached(used + required > limit).
> Reference codes in RegularContainerAllocator#assignContainer:
> {code:java}
> // How much need to unreserve equals to:
> // max(required - headroom, amountNeedUnreserve)
> Resource headRoom = Resources.clone(currentResoureLimits.getHeadroom());
> Resource resourceNeedToUnReserve =
> Resources.max(rc, clusterResource,
> Resources.subtract(capability, headRoom),
> currentResoureLimits.getAmountNeededUnreserve());
> boolean needToUnreserve =
> Resources.greaterThan(rc, clusterResource,
> resourceNeedToUnReserve, Resources.none());
> {code}
> For example, resourceNeedToUnReserve can be <8GB, -6 cores, 0 gpu> when
> {{headRoom=<0GB, 8 vcores, 0 gpu>}} and {{capacity=<8GB, 2 vcores, 0 gpu>}},
> needToUnreserve which is the result of {{Resources#greaterThan}} will be
> {{false}}. This is not reasonable because required resource did exceed the
> headroom and unreserve is needed.
> After that, when reaching the unreserve process in
> RegularContainerAllocator#assignContainer, unreserve process will be skipped
> when shouldAllocOrReserveNewContainer is true (when required containers >
> reserved containers) and needToUnreserve is wrongly calculated to be false:
> {code:java}
> if (availableContainers > 0) {
> if (rmContainer == null && reservationsContinueLooking
> && node.getLabels().isEmpty()) {
> // unreserve process can be wrongly skipped when
> shouldAllocOrReserveNewContainer=true and needToUnreserve=false but required
> resource did exceed the headroom
> if (!shouldAllocOrReserveNewContainer || needToUnreserve) {
> ...
> }
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
