must restart grpcui when connection to server is lost about grpcui HOT 3 CLOSED

This still seems to be an issue if the server panics or if the connection isn't shut down correctly.
Not always, but most of the time... Not sure yet as of the exact circumstances.

An example error is:

Unavailable (14)
all SubConns are in TransientFailure, latest connection error: connection error: desc = "transport: error while dialing: dial tcp [::1]:8761: connectex: No connection could be made because the target machine actively refused it."

The server is restarted and up and running, a restart of grpcui fixes it

from grpcui.

@PatrLind, that appears to be a different issue due to this part of the error message:

No connection could be made because the target machine actively refused it.

I suspect that is caused by a failure to reconnect when the server was down, and then RPC client using exponential backoff before re-trying to connect. So you may have caught it after its connection attempt was refused but before it re-tried the connection.

So I don't think grpcui needed to be restarted -- I think perhaps you needed to wait a little longer for the connection to heal. If you can try to verify that and if that's the case (it does eventually heal), file it as a different bug. If it never heals, try using -vv logging options and include the log output.

from grpcui.

I can confirm that it is indeed caused by the exponential backoff.
I understand that grpcui uses the default grpc backoff settings (max 2 minutes), right?
Would it be possible to either change this to maybe a few seconds by default, or perhaps (better) add a configuration flag to set the BackoffConfig.MaxDelay?

from grpcui.