It is helpful to have metrics about the number of jobs processed, or the speed of the jobs. But, sometimes, we may also want to know whether the machine that is hosting the Runner is experiencing operational issues, such as low memory or CPU contention. The Prometheus exporter also records general process metrics such as these.
General process metrics
Key metrics to watch
The following metrics only cover the host in which the Docker Machine executor is running:
- process_cpu_seconds_total: If you have a small machine, it is useful to know whether the CPU is coping with the load. There is a count that records the total number of CPU seconds used by the Runner process:
A big increase in a short period of time could indicate...
