This message was deleted.

When I understand correctly, Rancher Desktop takes the proxy configuration from the environment variables

HTTP_PROXY

,

HTTPS_PROXY

and

NO_PROXY

. This is all set in my environment. I use

px

proxy (https://github.com/genotrance/px) as proxy and have set the variables

HTTP_PROXY

and

HTTPS_PROXY

to

<http://localhost:3128>

. However, I still get errors in the

k3s.log

:

failing or missing response from https://<some ip>:10250/apis/metrics.k8s.io/v1beta1: Get "https://<some ip>:10250/apis/metrics.k8s.io/v1beta1": proxy error from 127.0.0.1:6443 while dialing <some ip>:10250, code 503: 503 Service Unavailable

well for the RD VM you don’t want K8s traffic to be forwarded to the proxy. That’s been a sore spot in the product as if you inject a proxy variable it can cause other issues in K8s since cluster traffic may be forwarded and not just calls to a internet registry or update channel call. The Electron side of RD uses your OS proxy settings.

Guess I am stuck with Docker Desktop - this works fine.

you can poke around https://github.com/rancher-sandbox/rancher-desktop/labels/area%2Fproxy to see what folks have tried. Tradeoffs in every approach.

Too bad. I would prefer Rancher Desktop over Docker Desktop. There must be a way. I mean it works for Docker Desktop with Kubernetes so it should with Rancher Desktop.

DD uses an internal proxy where the forwarding rules are handled. Just a significant architectural change compared to managing env vars

Alright... 😕

yup am in the same place. i find minimal issues on macos using the override.yaml, but it’s the WSL that I have not been able to handle consistently.

FYI there's a #rancher-desktop channel

Yes, I saw that but a little to late. This is why I have forwarded it to the #rancher-desktop channel.