I get a 503 error right after creating a workspace.

[kellyhyomin]

Summary

• che version : 7.3.0

• runtime : kubernetes

• screenshots

• my test devfile.yaml

  
  ---
  apiVersion: 1.0.0
  metadata:
  generateName: cdr-
  components:

• type: cheEditor id: cdr/code-server/2.1523-vsc1.38.1 memoryLimit: 1024Mi

I registered the code-server(Editor) with devfile-registry and added a stack. I get a 503 error right after creating a workspace on the stack. But when I refresh the browser, I get a normal vscode screen. I wonder if there is a way to fix it.

Relevant information

[benoitf]

@skabashnyuk liveness issue ? https://github.com/eclipse/che/blob/e3bc0a0bdab5df5b0da7ed49fc0371ee2e015852/assembly/assembly-wsmaster-war/src/main/webapp/WEB-INF/classes/che/che.properties#L158

[skabashnyuk]

That is by design. We need to know endpoint name to ping that has to be running for any case. In our case, this is cheEditor endpoint name. The predefined list you can find in che.workspace.server.liveness_probes variable. To handle that problem you have two options.

• adjust che.workspace.server.liveness_probes or CHE_WORKSPACE_SERVER_LIVENESS_PROBES
• or rename you endpoint name to match predefined che.workspace.server.liveness_probes=wsagent/http,exec-agent/http,terminal,theia,jupyter,dirigible https://github.com/eclipse/che-plugin-registry/blob/master/v3/plugins/dirigiblelabs/dirigible/3.4.0/meta.yaml#L15

[benoitf]

@skabashnyuk but code-server is provided by default on che so AFAIK che.properties should be updated according to all editors provided in Eclipse Che like https://github.com/eclipse/che-plugin-registry/blob/master/v3/plugins/cdr/code-server/2.1523-vsc1.38.1/meta.yaml#L15

[skabashnyuk]

Yes. Look so. @sr229 @kellyhyomin can you offer a pr?

[sr229]

I've been meaning to move it to the official images since my images were on testing phase, but officially this should have been fixed by some PRs in the che-plugin-registry already @sachinaraballi

[ghost]

@skabashnyuk I wonder that If I create a new alternative IDE in che like this link, should I use only the endpoint names in this list [wsagent/http, exec-agent/http, terminal, theia, jupyter, dirigible]? I think there should be an other endpoint name for the alternate IDEs. WDYT?

[skabashnyuk]

I think there should be an other endpoint name for the alternate IDEs.

If you don't want to use already defined names you have to adjust che.workspace.server.liveness_probes or CHE_WORKSPACE_SERVER_LIVENESS_PROBES. We may add some neutral or generic names like ide-server to defaults if you want.

[kellyhyomin]

@skabashnyuk @wonjooncho I added the 'ide-server' you told me in the configmap.yaml file and restarted che.

che/deploy/kubernetes/helm/che/templates/configmap.yaml

apiVersion: v1
kind: ConfigMap
metadata:
  labels:
    app: che
    component: che
  name: che
data:
  CHE_HOST: {{ template "cheHost" . }}
  CHE_PORT: "8080"
  .
  .
  CHE_WORKSPACE_SERVER_LIVENESS_PROBES: wsagent/http,exec-agent/http,terminal,theia,jupyter,dirigible,ide-server
.
.
.

my new alternative IDE meta.yaml

apiVersion: v2
publisher: popcornsar
name: webio
version: latest
type: Che Editor
displayName: AutoSAR Webio R19-03-V3
title: AutoSAR Webio R19-03
description: AutoSAR Webio
icon: https://www.eclipse.org/che/images/logo-eclipseche.svg
category: Editor
repository: https://github.com/eclipse/che
firstPublicationDate: '2019-02-05'
spec:
  endpoints:
  - name: ide-server
    public: true
    targetPort: 8080
    attributes:
      protocol: http
      type: ide
  containers:
  - name: webio
.
.
.

I specified ide-server as the endpoint name but I also faced a 503 error. There is no 503 error when changing the endpoint name to 'theia' you provide. I wonder if other settings are needed when using ide-server or other custom names.

[skabashnyuk]

@kellyhyomin can you provide a devfile that I can test your setup?

[che-bot]

Issues go stale after 180 days of inactivity. lifecycle/stale issues rot after an additional 7 days of inactivity and eventually close.

Mark the issue as fresh with /remove-lifecycle stale in a new comment.

If this issue is safe to close now please do so.

Moderators: Add lifecycle/frozen label to avoid stale mode.