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

[BUG] kube-ovn-cni consums lots of CPU when there are lots of subnets #4859

Open
oilbeater opened this issue Dec 20, 2024 · 0 comments · May be fixed by #4860
Open

[BUG] kube-ovn-cni consums lots of CPU when there are lots of subnets #4859

oilbeater opened this issue Dec 20, 2024 · 0 comments · May be fixed by #4860
Labels
bug Something isn't working performance Anything that can make Kube-OVN faster subnet

Comments

@oilbeater
Copy link
Collaborator

Kube-OVN Version

v1.13.2 and before

Kubernetes Version

Client Version: v1.28.2
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.32.0

Operation-system/Kernel Version

"Ubuntu 22.04.4 LTS"

Description

Related to #4849 and #4822 , when lots of subnets in a cluster, kube-ovn-cni can consume up to 80% of CPU.

This is related to the setIptables function which create count(subnet)*2 forward rules and check if rule exists every 10 seconds, which leads to a count(subnet)2count(subnet)*2 times filter.

Steps To Reproduce

  1. Create a cluster with 2 node and 250 subnets.
  2. Use top to see kube-ovn-daemon cpu usages

Current Behavior

Up to 80% CPU usage.

Expected Behavior

The CPU usage should be low under this load.

@oilbeater oilbeater added bug Something isn't working performance Anything that can make Kube-OVN faster labels Dec 20, 2024
@dosubot dosubot bot added the subnet label Dec 20, 2024
@oilbeater oilbeater linked a pull request Dec 20, 2024 that will close this issue
1 task
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working performance Anything that can make Kube-OVN faster subnet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant