Ingress NGINX Pod in 'Pending' State

[thepaulmacca]

What happened:

When deploying the Ingress NGINX controller to a new cluster, the pod is in a Pending state

What you expected to happen:

Pod to be in a Running state

How to reproduce it (as minimally and precisely as possible):

kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/main/deploy/static/provider/kind/deploy.yaml

Anything else we need to know?:

This only happens with the v0.13.0 release, it works fine in v0.12.0 so have just reverted to that for now

Environment:

• kind version: (use kind version): v0.13.0
• Kubernetes version: (use kubectl version): v1.24.0
• Docker version: (use docker info): 20.10.14
• OS (e.g. from /etc/os-release): WSL2 (Windows 10)

[BenTheElder]

I'm pretty sure this is ingress-nginx not being ready for Kubernetes 1.24 kubernetes/ingress-nginx#8091

[BenTheElder]

If you use kind create cluster --image=kindest/node:v1.23.6@sha256:1af0f1bee4c3c0fe9b07de5e5d3fafeb2eec7b4e1b268ae89fcab96ec67e8355 to switch to Kubernetes 1.23.6 I suspect it will work with kind v0.13.0

[thepaulmacca]

Ah right ok, I'll try that and report back if any issues. Thanks for the quick reply

[BenTheElder]

we may need to add a note to our docs in the meantime, if that is the case. https://kind.sigs.k8s.io/docs/user/ingress/
testing this is also on my TODO now ...

[BenTheElder]

Perhaps you just needed to wait longer for the pod to become ready?

I went back through https://kind.sigs.k8s.io/docs/user/ingress/ with kind v0.13.0 (default image) and successfully curled the foo and bar test services by the end.

One of the steps after installing the ingress is:

kubectl wait --namespace ingress-nginx \
  --for=condition=ready pod \
  --selector=app.kubernetes.io/component=controller \
  --timeout=90s

The pods can take a bit to settle.

[Alainx277]

I've experienced this same error on version 0.13.0.

I think it's caused by the new kubernetes version renaming the taint from node-role.kubernetes.io/master to node-role.kubernetes.io/control-plane. In the deploy.yaml file linked in the docs, it only tolerates the old taint name: https://github.com/kubernetes/ingress-nginx/blob/main/deploy/static/provider/kind/deploy.yaml#L479-L481

@BenTheElder Can we reopen this issue?

[BenTheElder]

We can, but this may not be the same issue, for single node clusters we already remove the taint.
I suspect your case is different than OP and the actual change needs to happen in ingress-nginx repo.

[BenTheElder]

Since no issues were reported back, I would prefer a new bug with the bug template filled out with your reproducer etc.

[BenTheElder]

Upstream: kubernetes/ingress-nginx#8605

[BenTheElder]

kubernetes/ingress-nginx#8609 is merged upstream