Skip to content
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.

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

[Improvement] ShuffleServer set MALLOC_ARENA_MAX #859

Closed
3 tasks done
rhh777 opened this issue May 10, 2023 · 0 comments · Fixed by #860
Closed
3 tasks done

[Improvement] ShuffleServer set MALLOC_ARENA_MAX #859

rhh777 opened this issue May 10, 2023 · 0 comments · Fixed by #860
Assignees

Comments

@rhh777
Copy link
Contributor

rhh777 commented May 10, 2023

Code of Conduct

Search before asking

  • I have searched in the issues and found no similar issues.

What would you like to be improved?

When the shuffle server is deployed in the container, the memory keeps soaring until it exceeds the container's memory limit, and then the container is OOMKill. After checking the memory usage, it is found that there are a large number of 64 MB memory blocks.
Reference:
prestodb/presto#8993
https://issues.apache.org/jira/browse/HADOOP-7154
image

image

How should we improve?

In start-shuffle-server.sh, set MALLOC_ARENA_MAX(default cpu core * 8) to control memory application/usage

Are you willing to submit PR?

  • Yes I am willing to submit a PR!
rhh777 added a commit to rhh777/incubator-uniffle that referenced this issue May 10, 2023
jerqi pushed a commit that referenced this issue May 10, 2023
…860)

### What changes were proposed in this pull request?

Reduce memory usage for some versions of glibc. Prevent oom.

### Why are the changes needed?

Fix: #859

### Does this PR introduce _any_ user-facing change?

No.

### How was this patch tested?
UT

Co-authored-by: haorenhui <haorenhui@kingsoft.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant