Failed to configure Gitlab Workspace

Hello everyone!
I would like to ask for support regarding the configuration of Gitlab Workspace (Beta).
This is an interesting feature for me and I would like to use it for some projects.
I followed the configuration steps Gitlab Workspace - Doc. When I create a new Workspace, I get Failed after 10 min.

In my configuration, I think the ingress-nginx is the problem. I couldn’t find a good example to apply. Here’s the code

kind: Ingress
  name: my-ingress
    - host: my-dns-record
          - path: /
            pathType: Prefix
                name: my-ingress-ingress-nginx-controller
                  number: 80

I’m using cloudflare for DNS record creation and TXT type (used by certbot)

My agent yaml file

  enabled: true
  dns_zone: "my-dns-record"

and my DevFile

schemaVersion: 2.2.0
  - name: tooling-container
      gl/inject-editor: true
      memoryRequest: 1024M
      memoryLimit: 2048M
      cpuRequest: 500m
      cpuLimit: 1000m
      - name: http-3000
        targetPort: 3000

The proxy was successfully installed ~ GitLab Workspaces Proxy is now installed in the cluster.
Please let me know if you need any additional information.
Thank you in advance

Hey @karimchaouch

There’s a few more pieces to a ‘complete’ ingress configuration than just the Ingress object.

The easiest way to install ingress-nginx is to use helm - that takes care of all the working pieces.

This worked for me:

helm repo add ingress-nginx
helm repo update
helm install nginx-ingress ingress-nginx/ingress-nginx --namespace ingress-nginx --create-namespace

Once that’s complete, you should have some pods/services running in the ingress-nginx namespace - and then after the gitlab-workspaces-proxy you should have the right entries if you run kubectl get ingress -A

1 Like

After checking more in details and fixing issues related to pvc. I can see the pod getting created. But still failing after few seconds…

here’s the log output:

kubectl logs workspace-71591-15054274-ubvxsu-77785d7df6-b26wl -n gl-rd-ns-71591-15054274-ubvxsu -v=4

Defaulted container "tooling-container" out of: tooling-container, gl-cloner-injector-gl-cloner-injector-command-1 (init), gl-editor-injector-gl-editor-injector-command-2 (init)
/projects/.gl-editor/code-server/bin/gitlab-webide-server: line 12: /projects/.gl-editor/code-server/node: not found

I fixed the issue by creating a persistent storage and claiming it.
That’s why I was stuck at creating for 10 min. Hopefully that will help someone else.

I was getting these errors pod has unbound immediate persistentvolumeclaims

Great, glad you managed to get it working :slight_smile:

This is currently mentioned in the prerequisites:

In the Kubernetes cluster, verify that a default storage class is defined so that volumes can be dynamically provisioned for each workspace.

Our documentation is open source, and we’d love an MR if you think we could make that particular call out any clearer or easier to follow for someone else…

1 Like

Thanks @stingrayza.
Yes, it would be nice to mention the persistent storage in the prerequisites. I will create a MR


1 Like