flannel: Flannel POD OOMKilled

Flannels pod have 600 restart on node only (not on master) due to OOMKilled.

Expected Behavior

No Restart

Possible Solution

Increase memory limit

Steps to Reproduce (for bugs)

install flannel using prodived yaml (Documentation/kube-flannel.yml)

resources:
          requests:
            cpu: "100m"
            memory: "50Mi"
          limits:
            cpu: "100m"
            memory: "50Mi"

Context

1 Master / 2 Node

Your Environment

Flannel version: v0.10.0-amd64
Backend used (e.g. vxlan or udp): vxlan
Etcd version: 3.1.11
Kubernetes version (if used): 1.9.4
Operating System and version: Ubuntu 16.04.4

What’s the recommended memory limit for Flannel ?

About this issue

Original URL
State: closed
Created 6 years ago
Comments: 16 (2 by maintainers)

Most upvoted comments

We have updated resources limits and there are no more restart. So the question is :

What's the recommended memory limit for Flannel running in Kubernetes node's pod ?

jbpin on Mar 22, 2018

The same issue for me. Fixed by setting 256Mi memory limit in flannel daemonSet properties

Tapolsky on Sep 5, 2018

@shettyh

Maybe you can modify the memory limit, and test it. Try this command:

kubectl patch ds -n=kube-system kube-flannel-ds-amd64 -p ‘{“spec”: {“template”:{“spec”:{“containers”: [{“name”:“kube-flannel”, “resources”: {“limits”: {“cpu”: “250m”,“memory”: “550Mi”},“requests”: {“cpu”: “100m”,“memory”: “100Mi”}}}]}}}}’

lubinsz on Jul 4, 2019