Skip to content

YARN-11843. Fix potential deadlock when auto-correction of container allocation is enabled #7855

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: trunk
Choose a base branch
from
Open

YARN-11843. Fix potential deadlock when auto-correction of container allocation is enabled #7855

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
b8c0647
YARN-11843. Fix potential deadlock when auto-correction of container …
TaoYang526 Aug 5, 2025
cc6bc5f
Update comment.
TaoYang526 Aug 5, 2025
29cd1d7
Fix checkstyle warning.
TaoYang526 Aug 5, 2025
f05899c
Adjust expected state from EXIPRED to RELEASED in AbstractYarnSchedul…
TaoYang526 Aug 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 1 addition & 3 deletions ...n/java/org/apache/hadoop/yarn/server/resourcemanager/scheduler/AbstractYarnScheduler.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -724,9 +724,7 @@ protected void autoCorrectContainerAllocation(List<ResourceRequest> resourceRequ
if (extraContainers > 0) {
// Change the state of the container from ALLOCATED to EXPIRED since it is not required.
LOG.debug("Removing extra container:{}", rmContainer.getContainer());
completedContainer(rmContainer, SchedulerUtils.createAbnormalContainerStatus(
rmContainer.getContainerId(), SchedulerUtils.EXPIRED_CONTAINER),
RMContainerEventType.EXPIRE);
asyncContainerRelease(rmContainer);
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Will it cause the message at the end of the container to be different?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, the message will be updated from "Container expired since it was unused" to "Container released by application", which I believe is more appropriate.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, the message will be updated from "Container expired since it was unused" to "Container released by application", which I believe is more appropriate.

No, I think the old is more clear.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For the old one, "expired since it was unused" is not accurate because it was released due to over-allocation, not because the waiting period expired. For the new one, "Container released by application" has been used in different scenarios that actively released by application, I think over-allocation could be included in this scope. @shameersss1 could you please share your thoughts about this?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@TaoYang526 +1
Container released makes more sense here.

application.newlyAllocatedContainers.remove(rmContainer);
extraContainers--;
}
Expand Down