Errors until broken pipe after ssh into my devpod

[rafilkmp3]

Errors until broken pipe after ssh into my devpod

[pascalbreuninger]

Yup we've heard this issue from other people as well, will fix as soon as possible! A way to work around it for now is to use prebuilds from your local machine and then pull the prebuild image inside of your cluster, removing the need to build the features in-cluster

edit: @rafilkmp3 Seems like I wrote a bit too hastily, @89luca89 already fixed that one a while ago 🥳
Can you try again with one of the 0.5.0 alpha preleases? We're planning to cut a new stable release end january

[pascalbreuninger]

@rafilkmp3 thank you for reporting the issue.
Could you copy the SSH config entry for $WORKSPACE_ID.devpod please?
Additionally, can you manually exec into the pod kubectl exec -n $NS $POD_NAME -c devpod bash?

[rafilkmp3]

Sure, thank you for the reply.
Here is my SSH config:

# DevPod Start k8s.devpod
Host k8s.devpod
  ForwardAgent yes
  LogLevel error
  StrictHostKeyChecking no
  UserKnownHostsFile /dev/null
  ProxyCommand /opt/homebrew/bin/devpod ssh --stdio --context default --user vscode k8s
  User vscode
# DevPod End k8s.devpod

I have tried manually exec, but no success,
kubectl exec -n devpod devpod-k8s-defaul-d931e -c devpod -- bash

[pascalbreuninger]

thanks @rafilkmp3
What's the error output of your kubectl command?

edit: whilst going through issues I realized there are flags missing in the command. Try
kubectl exec -it -n devpod devpod-k8s-defaul-d931e -c devpod -- bash

[rafilkmp3]

Hello @pascalbreuninger, sorry for the delayed reply, I've been debugging it recently and I was able to figure out the broken pipe issue.
However, I'm now stuck at something else, i can get my devpod up using my EKS with no problems whatsoever, but when adding a new "features" field, i get errors pointing at a failure finding the /tmp directory.

Here is my .devcontainer.json
{ "name": "DevPod Demo", "image": "ubuntu", "features": { // "ghcr.io/devcontainers/features/terraform:1": {} } }

and here are the debug logs:
11:13:31 info Use Kubernetes Namespace 'devpod' 11:13:31 info Use Kubernetes Config '/Users/rafael.sathler/.kube/config' 11:13:31 info Use Kubernetes Context 'arn:aws:eks:us-east-2:REDACTED' 11:13:31 info Run command: kubectl --namespace devpod --kubeconfig /Users/rafael.sathler/.kube/config --context arn:aws:eks:us-east-2:REDACTED' exec -c devpod -i devpod-k8s-defaul-b1630 -- sh -c '/usr/local/bin/devpod' helper ssh-server --stdio --debug 11:13:34 debug SSH client created 11:13:34 debug SSH session created 11:13:35 info Exit error: start agent: stat /tmp: no such file or directory 11:13:35 info start agent: stat /tmp: no such file or directory 11:13:35 debug Connection to SSH Server closed 11:13:35 debug Done setting up container 11:13:35 info Process exited with status 1 11:13:35 info run agent command 11:13:35 info github.com/loft-sh/devpod/pkg/devcontainer/sshtunnel.ExecuteCommand.func2 11:13:35 info /Users/runner/work/devpod/devpod/pkg/devcontainer/sshtunnel/sshtunnel.go:111 11:13:35 info runtime.goexit 11:13:35 info /Users/runner/hostedtoolcache/go/1.20.5/x64/src/runtime/asm_arm64.s:1172 11:13:35 info devcontainer up 11:13:35 info github.com/loft-sh/devpod/cmd/agent/workspace.(*UpCmd).Run 11:13:35 info /Users/runner/work/devpod/devpod/cmd/agent/workspace/up.go:95 11:13:35 info github.com/loft-sh/devpod/cmd/agent/workspace.NewUpCmd.func1 11:13:35 info /Users/runner/work/devpod/devpod/cmd/agent/workspace/up.go:52 11:13:35 info github.com/spf13/cobra.(*Command).execute 11:13:35 info /Users/runner/work/devpod/devpod/vendor/github.com/spf13/cobra/command.go:916 11:13:35 info github.com/spf13/cobra.(*Command).ExecuteC 11:13:35 info /Users/runner/work/devpod/devpod/vendor/github.com/spf13/cobra/command.go:1044 11:13:35 info github.com/spf13/cobra.(*Command).Execute 11:13:35 info /Users/runner/work/devpod/devpod/vendor/github.com/spf13/cobra/command.go:968 11:13:35 info github.com/loft-sh/devpod/cmd.Execute 11:13:35 info /Users/runner/work/devpod/devpod/cmd/root.go:90 11:13:35 info main.main 11:13:35 info /Users/runner/work/devpod/devpod/main.go:8 11:13:35 info runtime.main 11:13:35 info /Users/runner/hostedtoolcache/go/1.20.5/x64/src/runtime/proc.go:250 11:13:35 info runtime.goexit 11:13:35 info /Users/runner/hostedtoolcache/go/1.20.5/x64/src/runtime/asm_arm64.s:1172 11:13:35 debug Connection to SSH Server closed 11:13:35 debug Done creating devcontainer 11:13:35 debug Done executing ssh server helper command 11:13:35 fatal Process exited with status 1 run agent command github.com/loft-sh/devpod/pkg/devcontainer/sshtunnel.ExecuteCommand.func2 /Users/runner/work/devpod/devpod/pkg/devcontainer/sshtunnel/sshtunnel.go:111 runtime.goexit /Users/runner/hostedtoolcache/go/1.20.5/x64/src/runtime/asm_arm64.s:1172

I've tried multiple features but all of them seems to lead to the same error.
I am using a multiarch cluster, maybe could it work using node selector and forcing the usage of AMD64?