Error Starting Kubernetes on Windows

Actual Behavior

After installing 1.3.0 I tried starting kubernetes, but it fails.

Steps to Reproduce

• Install Rancher Desktop
• Start Kubernetes (v1.19.16 w/o Traefik either containerd and dockerd)

Result

Error Starting Kubernetes

Error: C:\Users\xxx\AppData\Local\Programs\Rancher Desktop\resources\resources\win32\bin\kubectl.exe exited with code 1

Last command run: C:\Users\xxx\AppData\Local\Programs\Rancher Desktop\resources\resources\win32\bin\kubectl.exe config current-context

Context: Waiting for services

 stdout: ''
}
2022-04-29T07:45:55.262Z: WSL: executing: readlink -f $HOME/.docker/cli-plugins/docker-compose: Error: wsl.exe exited with code 1
2022-04-29T07:45:55.262Z: Failed to readlink/rm $HOME/.docker/cli-plugins/docker-compose Error: wsl.exe exited with code 1
    at ChildProcess.<anonymous> (C:\Users\xxx\AppData\Local\Programs\Rancher Desktop\resources\app.asar\dist\app\background.js:17:141690)
    at ChildProcess.emit (node:events:390:28)
    at Process.ChildProcess._handle.onexit (node:internal/child_process:290:12) {
  stdout: ''
}

Expected Behavior

Kubernetes Starts

Additional Information

No response

Rancher Desktop Version

1.3.0

Rancher Desktop K8s Version

1.19.16

Which container runtime are you using?

containerd (nerdctl)

What operating system are you using?

Windows

Operating System / Build Version

Windows 10 21H2 (Build 19044.1645)

What CPU architecture are you using?

x64

Linux only: what package format did you use to install Rancher Desktop?

No response

Windows User Only

• VPN, but only for company internal traffic, wich is not the case for RD
• Trendmicro

[jandubois]

Can you give me a hint what the issue is so maybe I can fix my setup to make it work?

I can’t tell from the log snippets what the real issue is. It looks like it is related to file system mounts, so maybe a WSL2 configuration problem? Given that things work fine for most users, I suspect it will be something you could fix yourself, but we first need to understand why it is failing.

It would help to get the full set of logs and not just the last couple of lines. Could you zip them up and attach here?

am I the only one who has this issue?

Hard to tell, there may be one or maybe 2 other issues that may be related, but it isn’t clear. The vast majority of the users do not have this problem, and it did not come up in our manual testing either.

Why is it moved to “Later”, while it actually makes Rancher Desktop unusable for everyone who encounters this issue…

The 1.4.0 release is essentially done; only high-severity fixes will still make it in, and problems found during manual QA.

If we wait until there are no open issues left, then there won’t be another release ever again… If we had lots of reports of this particular breakage, then it would be considered a blocker, but anything affecting only a few users, especially if the cause of the problem is still unknown, cannot stop the release.

I’m sorry this is not working for you, and I hope you can provide additional information that will help us to track this down.

[jandubois]

I’m going to re-open this issue, but don’t have any suggestion how to recover RD from running it as admin before. We will need #2186 implemented, so you can run once as admin, perform “Factory Reset” to remove all items owned by the admin account, and then start the app again as a regular user. This will hopefully be implemented in the 1.5.0 release.

Right now Rancher Desktop will restart automatically after a Factory Reset, so all files/directories will be recreated by the admin user again.

Furthermore we have #1560 that should prevent a user from getting into the situation in the first place.