Closed jinnatar closed 2 years ago
Add a Prometheus compatible metrics endpoint to daemon mode:
--metrics
:8080
--listen-address
Instrument Scheduler and job runner. Metrics provided are:
The most complex metric is chadburn_run_duration_seconds since it's a histogram. Example PromQL query to get average runtimes with a 5 minute window:
rate(chadburn_run_duration_seconds_sum[5m]) / rate(chadburn_run_duration_seconds_count[5m])
Sample graph from that query on 3 relatively frequent and fast jobs: https://prnt.sc/cp9DBKKoCQ1t
Add a Prometheus compatible metrics endpoint to daemon mode:
--metrics, not enabled by default.:8080, override with--listen-address.Instrument Scheduler and job runner. Metrics provided are:
The most complex metric is chadburn_run_duration_seconds since it's a histogram. Example PromQL query to get average runtimes with a 5 minute window:
Sample graph from that query on 3 relatively frequent and fast jobs: https://prnt.sc/cp9DBKKoCQ1t