We recently upgraded from 18.5 to 20.1. Following the upgrade we encountered hundreds of intermittent dropped connections between the server and the agents. This manifested itself in a couple of ways.
nested exception is javax.net.ssl.SSLHandshakeException: Remote host terminated the handshake
nested exception is org.apache.http.NoHttpResponseException: <hostname>:8154 failed to respond
I'm guessing this is due to load but at most we have 160 agents running and we didn't experience anything like this on 18.5. Due to this we had to downgrade back to 18.5.
I've looked at thread pool settings, timeouts, max connections etc... but nothing appears to have changed between versions. Is there other settings that I can look at to better tune this?