Closed wwadge closed 4 years ago
Perhaps earlier pings are throwing some exception that kills off the keep-alive thread. Will test and report back
Unfortunately trapping the exception, whilst probably is a good idea anyway, still doesn't resolve this issue. Might have to go down the route of opening a new connection if need be but I don't think this is the best approach
Opening a new connection worked for me. Still can't figure out why we get Timeouts on the same connection that has ongoing traffic. Perhaps this is something on Binance side not using "ping" as a signal to reset their idle timeout.
99% of these timeout errors were synchronization issues that I've fixed in a separate branch that I just merged into master. Should be OK now. Thanks a lot for your help!
When a signal is triggered, I often get the following exception. I put in an emergency exit (System.exit) upon repeated failures and restart the app at which point the connection works straight away. I haven't yet been able to track this one down yes -- I've already added a keep alive timer that also pings the connection. Leaving this bug report here until I can nail this one down (or perhaps open a new restClient() connection each time?)