Health Check - TCP RST - ErrorPolicyTrace

Staking & Delegation Operate a Stake Pool
1

Every 10s I have a health check probe that connects and then sends a TCP RST packet to verify that my nodes are online. As a result, I get many of the following warnings which are flooding my logs:

{"thread":"156","loc":null,"data":{"address":"xxx.xxx.xxx.xxx","kind":"ErrorPolicyTrace","event":"ErrorPolicySuspendPeer (Just (ApplicationExceptionTrace (MuxError MuxBearerClosed \"<socket: 42> closed when reading data, waiting on next header True\"))) 20s 20s"},"sev":"Warning","env":"1.29.0:4c594","msg":"","app":[],"host":"alonzo","pid":"12636","ns":["cardano.node.ErrorPolicy"],"at":"2021-09-19T09:49:07.81Z"}

Is there an ASCII value I can send when the probe connects so that it does not cause this exception/warning log?

The only other solution I can think of is to disable TraceErrorPolicy in the config, which I would prefer not to do. I cannot use minSeverity: "Error" either as that will prevent gLiveView from functioning which is very handy for checking peers, etc.

2

Hey Ryan, I think you may not be getting responses on this topic because it seems like you are monitoring your server with a tool that we are not familiar with and may not be a supported method.

I will say that most SPOs are using Prometheus/Node Exporter for data aggregation and display it on a Grafana Dashboard. You can even connect the Grafana Dashboard to a Telegram bot and get health notification/metrics/screenshots right to your phone. It really works well and natively supported by the node.

Hope this helps.