Watch API. Handle connection loss.

[bm0]

Hey.
I found that when the connection is disconnected, due to a restart of the etcd server or for other reasons, reading from the watch channel is blocked forever.

I am not an GRPC expert, maybe this problem can be solved if the connection is configured correctly (I'm talking about grpc.DialOption)

Is there a way to handle connection loss?
I would like to get an error from the watcher channel so that I can reconnect later when network or other problems are resolved. Help me please. If possible, then with an example. Thanks in advance.

[agargi]

@bm0 Looks like you are looking to something like this (refer link below). It basically wraps the context with 'WithRequireLeader' which ensures that a failure is returned when cluster is unable to determine a leader (after 3 re-election timeouts and each election timeout defaulting to 1000 ms).

kubernetes/kubernetes#89488

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed after 21 days if no further activity occurs. Thank you for your contributions.