Description of problem:

If the JH leader pod dies during the spawn process for a server, although the notebook server is spawned and running, the user will not be automatically redirected like it happened before.

The spawn dialog will fail like so:

Closing the dialog and trying to spawn again will result in the same error. Clicking on the "Try Again" button will instead send the user to their server, and the same thing happens when refreshing the page.

I believe this is a regression compared to earlier behaviour, where the user would get automatically redirected once the new leader was elected without needing to take action. If an automatic redirect is not possible with the new spawner, I believe this new behaviour should be at least documented for users.

Prerequisites (if any, like setup, operators/versions):

RHODS 1.5.0-2 on OSD (OCP 4.9)

Steps to Reproduce

1. Start server spawn with any image
2. Kill leader pod
3. Check running pods in rhods-notebooks

Actual results:

user server pod is running but spawner does not redirect the user automatically and displays an error message. Refreshing the page or clicking "Try again" in the modal redirects the user to the running server.

Expected results:

User should be automatically redirected to their server once the new leader pod is elected. Spawner should not be stuck showing an error message while the user server is actually running.

Reproducibility (Always/Intermittent/Only Once):

Always

Build Details:

RHODS 1.5.0-2 on OSD

Workaround:

Refresh page or click on "Try again" in spawner dialog.

Additional info: