Trying to debug a production issue and trying to locate where the issue is originating; I am using websockets with pedestal through the jetty websocket configurator, and I’m seeing a repeating log for :on-close firing with {:code 1006, :reason "WebSocket Read EOF"}. I’m not super familiar with low-level websocket details, and I’m wondering if someone can help me determine if this is the pedestal server’s issue, or if it’s just closing the connection in response to a problem from the other side?

@dadair do you have a ping method on your websocket? I dont remember if it's the 1006, but I fixed some "EOF error" doing a ping every minute

@dadair the pedestal user group has broader reach. You can try asking there as well.

@souenzzo we have a ping mechanism set up, the issue appears to be related somehow to our nginx proxy “degrading” when deployed in kubenetes; seems a proxy reset fixed the issue /shrug