Kind cluster not able to pull public images from ghcr.io

rohitagg2020 commented 2 weeks ago

What happened:

I have created a kind cluster. But when I try to create a deployment (which pulls images from ghcr.io), it fails.

Error:

^ Pending: ErrImagePull (message: failed to pull and unpack image "ghcr.io/carvel-dev/kapp-controller@sha256:b83bd139c42777bb4428c2ae421f93675a7d6147dff3147471d1be112d9c1eec": 
failed to resolve reference "ghcr.io/carvel-dev/kapp-controller@sha256:b83bd139c42777bb4428c2ae421f93675a7d6147dff3147471d1be112d9c1eec": 
failed to do request: Head "https://ghcr.io/v2/carvel-dev/kapp-controller/manifests/sha256:b83bd139c42777bb4428c2ae421f93675a7d6147dff3147471d1be112d9c1eec": 
tls: failed to verify certificate: x509: certificate signed by unknown authority)

It has been working fine until last month. I thnk I have messed up my environment, because of which I am facing this issue.

Things I have tried:

Uninstall and install kind.
Deleted and created new kind cluster.
Uninstalled and reinstalled docker desktop.

But still, I am not able to get around the error. How can I get back to things working again?

What you expected to happen: Deployment to be created successfully. How to reproduce it (as minimally and precisely as possible): kubectl apply -f https://github.com/carvel-dev/kapp-controller/releases/latest/download/release.yml Anything else we need to know?: As this is a public image, I guess the ca certs are not being picked up when kind cluster is created. Environment:

kind version: (use kind version): kind v0.23.0
Runtime info: (use docker info, podman info or nerdctl info): Client: Docker Engine - Community Version: 26.1.4 Context: desktop-linux Debug Mode: false
OS (e.g. from /etc/os-release): macOS
Kubernetes version: (use kubectl version): Client Version: v1.29.2 Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3 Server Version: v1.27.9
Any proxies or other special environment settings?:

stmcginnis commented 2 weeks ago

/remove-kind bug /kind support

Please fill in the full "Environment" details from the issue template.

Are you in an environment that uses a proxy?

You could try deleting any cluster you have, then delete the kind Docker network, then try creating again. It will try to recreate the network to make sure there isn't something odd with the existing setup.

You can also do kind create cluster --retain, then use docker ps to see the created containers. Then you can docker exec into the kind container(s) to perform troubleshooting.

BenTheElder commented 2 weeks ago

This looks like a MITM proxy causing issues.

rohitagg2020 commented 1 week ago

Are you in an environment that uses a proxy?

No

You could try deleting any cluster you have, then delete the kind Docker network, then try creating again. It will try to recreate the network to make sure there isn't something odd with the existing setup.