How does workspace idling timeout work in Eclipse Che with DW enabled?

[SkorikSergey]

Summary

How does workspace idling timeout work in Eclipse Che with DW enabled?

I tried to use next DevWorkspaceOperatorConfig object:

kind: DevWorkspaceOperatorConfig
apiVersion: controller.devfile.io/v1alpha1
metadata:
  name: devworkspace-operator-config
  namespace: openshift-operators
config:
  workspace:
    imagePullPolicy: Always
    idleTimeout: '1m'

After this object creation devworkspace-controller-manager pod updated configuration

{"level":"info","ts":1649081288.4094315,"logger":"operator-configuration","msg":"Updated config to [(default config)]"}
{"level":"info","ts":1649081310.787388,"logger":"operator-configuration","msg":"Updated config to [workspace.idleTimeout=3m]"} 

But workspace didn't stop after 3 min

{"level":"info","ts":1649084814.4158278,"logger":"controllers.DevWorkspace","msg":"Got conflict when trying to apply timing annotations to workspace","Request.Namespace":"opentlc-mgr-codeready","Request.Name":"cpp","devworkspace_id":"workspacef48ff8969efa4895"}
{"level":"info","ts":1649084814.4289684,"logger":"controllers.DevWorkspace","msg":"Error updating workspace status: Operation cannot be fulfilled on devworkspaces.workspace.devfile.io \"cpp\": the object has been modified; please apply your changes to the latest version and try again","Request.Namespace":"opentlc-mgr-codeready","Request.Name":"cpp","devworkspace_id":"workspacef48ff8969efa4895"}

Relevant information

Checked on Eclipse Che 7.45.0 on Openshift 4.9 deployed from OperatorHub.

[musienko-maxim]

It is interesting - when I run the WebTerminal instance which also creates DevWorkspace under the hood this feature works as expected. The WebTerminal has been stopped according to the configuration

[dmytro-ndp]

@musienko-maxim: Web terminal idle timeout had been decoupled from DevWorkspace idle timeout 26 days ago: redhat-developer/web-terminal-operator#114

[l0rd]

In Che the timeout is set in the definition of the editor. As a workaround you can point to an alternative editor definition (c.f. <che-URL>/dashboard/#https://github.com/che-samples/bash/tree/devfilev2?che-editor=https://gist.githubusercontent.com/vinokurig/8240d845067c89e295e2cb4dc2c0070b/raw/ad3452ee49b0eeddfaf086f55557d46206073fea/che-theia.yaml).

I will open an other issue to address that:

1. che admin should be able to specify default idling timeout via CheCluster CR
2. che-theia definition in the plugin regsitry should not specify machine exec idle timeout

[l0rd]

I have created #21390

[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.