buggood first issue
Repository-Metriken
- Stars
- (2.450 Sterne)
- PR-Merge-Metriken
- (Durchschn. Merge 10T 13h) (29 gemergte PRs in 30 T)
Beschreibung
What happened:
Warning FailedCreatePodSandBox 2m6s (x62 over 15m) kubelet (combined from similar events): Failed to create pod sandbox: rpc error: code = Unknown desc = failed to setup network for sandbox "f44b32fc51168ef2352e384a3d853da635c5fa0b469121f1aede7a924004b870": plugin type="aws-cni" name="aws-cni" failed (add): add cmd: failed to assign an IP address to container
This code notices unmanaged ENIs and increments the count, but never decrements it: https://github.com/aws/amazon-vpc-cni-k8s/blob/a2c18c1cb0ff32c36157b6dae4aacc5d5f243b3d/pkg/ipamd/ipamd.go#L2104
In our workflow we attach and detach unmanaged ENIs from our nodes for during some operations. Doing so can result in the aws-cni eventually failing to assign IPs to containers.
Attach logs
What you expected to happen: Detaching an unmanaged ENI should decrement the count.
How to reproduce it (as minimally and precisely as possible):
Anything else we need to know?:
Environment:
- Kubernetes version (use
kubectl version): - CNI Version
- OS (e.g:
cat /etc/os-release): - Kernel (e.g.
uname -a):