Right now, the current behavior for reconnection to sockets will cause an infinite loop of retrying connections if the server goes down. To mitigate issues like this, our providers should support some sort of strategy that uses a combination of
maximum retry connection count
backoffs to delay requests to re-connect using some strategy (linear or exponential)
alex-hourglass
commented
7 months ago
Right now, the current behavior for reconnection to sockets will cause an infinite loop of retrying connections if the server goes down. To mitigate issues like this, our providers should support some sort of strategy that uses a combination of