Program should stop retrying port-forwarding on Exit Code 1

[pophilpo]

The current implementation retries port-forwarding regardless of the exit code returned by the kubectl port-forward command. This behavior leads to unnecessary retries, especially when the exit code is 1, which indicates that the kubectl command itself was successful but the port-forwarding operation failed due to permissions or other non-retryable reasons.

If the exit code for the kubectl port-forward command is 1, the program should stop retrying and possibly alert the user, as this is typically a non-recoverable issue.

[pophilpo]

@sunsided do we really want to handle these separately ? Again, I feel like these are non-recoverable errors that we can just ignore with a message.

TODO: Handle `Error from server (NotFound): pods "foo-78b4c5d554-6z55j" not found")`
TODO: Handle `Unable to listen on port 5012: Listeners failed to create with the following errors: [unable to create listener: Error listen tcp4 127.1.0.1:5012: bind: address already in use]`

Currently my PR handles these as well, while also providing a descriptive error message

Spawning child processes:
#2: Unable to listen on port 5010: Listeners failed to create with the following errors: [unable to create listener: Error listen tcp4 127.0.0.1:5010: bind: address already in use]
#2: error: unable to listen on any of the requested ports: [{5010 80}]
#2: Process exited with exit status: 1 - shutting process down

[sunsided]

The core issue is that of forwarding to a service, not a pod directly. If all the available pods of a service disappear temporarily, say because they are recreated, then we do want to keep retrying until they're back. If we can guarantee that kubectl does not exit with code 1 (or exits with a different nonzero) when it could indeed recover from this situation, we're good. Otherwise we may want to introduce an option / configuration --max-retries=N or similar.