Local host connection refused issue with only one of many otherwise working alerts

I am running Grafana and InfluxDB on the same server, and have a number of metrics providing alerts without issue. However, one of the alerts fairly consistently emails me the following [Alerting] error immediately followed by an [OK] email.

tsdb.HandleRequest() error Get http://localhost:8086/query?db=solar&epoch=s&q=SELECT+sum%28%22power%22%29+-+sum%28%22ac_power%22%29+FROM+%22autogen%22.%22helioscope%22+WHERE+time+%3E+now%28%29+-+720h+GROUP+BY+time%281d%29+fill%28none%29: dial tcp [127.0.0.1:8086](http://127.0.0.1:8086/): connect: connection refused

While it’s just an annoyance since it immediately succeeds, I want to fix the problem before attempting to scale the system. For now it’s easy enough just to turn off the “If execution error or timeout” alerting.

not sure what could be the cause for this network error. For some reason influxdb stops accepting the connection

Any recommendations on which Influx logs I should look for?