che
che copied to clipboard
eclipse-che
Getting Started workspace fails to start if there is exsiting workspace
Describe the bug
Getting Started workspace fails to start if there is exsiting workspace
Che version
7.50@latest
Steps to reproduce
Expected behavior
exising cpp workspace is started
Runtime
OpenShift
Screenshots
No response
Installation method
OperatorHub
Environment
Linux
Eclipse Che Logs
No response
Additional context
No response
Ok but we should not restart the workspace because it has the same name: it may have the same name but refer to another branch (i.e. you have just reviewed a PR) or even be a completely different git repository. How are we going to figure out if that's the same workspace or a different one?
Can we just create a new workspace but with another name (maybe adding some suffix)
Can we just create a new workspace but with another name (maybe adding some suffix)
Right but that's what generateName, in the DevWorkspace spec, is supposed to do. If the problem here is with samples I think we can use generateName instead of name when generating the DevWorkspace.
I think samples are easy because we start from a DevWorkspace.yaml. @ibuziuk the problem here is with a sample (i.e. if fails the second time you try a sample) right?
I can create a distinct issues for non-samples but in general I would ask the user if he wants to restart the existing workspace or replace it with the new one. To create multiple workspace using the same devfile it's recommended to leave the metadata.name empty in the devfile.yaml or use the new query parameter. But again, I think that's a distinct issue.
I can create a distinct issues for non-samples but in general I would ask the user if he wants to restart the existing workspace or replace it with the new one. To create multiple workspace using the same devfile it's recommended to leave the metadata.name empty in the devfile.yaml or use the new query parameter. But again, I think that's a distinct issue.
@akurinnoy I have changed the title here because it looks that the issue is related to samples only. If that's not the case let's discuss it because the solution is not clear.