Does anyone have any best practices for relay / firehose / event-stream consumption? I’m working through a relay consumer that looks like it “just stopped” receiving events last night.
I’m assuming its some issue on my end, and I’d like to have some reasonable circuit breakers in place. So far the list includes:
- If I stop receiving events on an open / active ws connection after 3 minutes then warn
- If I stop receiving events on an open / active ws connection after 5 minutes then disconnect and reconnect at the last known good cursor
Any suggestions?