Cabourotte exposes various metrics about itself and about healthchecks on the /metrics endpoint using the Prometheus format.
Here are some interesting metrics exposed:
exporters
exporter_duration_seconds_bucket: The duration for the healthchecks exporters. Labels are status, which indicates if healthchecks were exported successfully or not, and name which is the exporter name.exporter_duration_seconds_count: The number of events exported to exporters. The labels are the same than the previous metric.result_chan_size: Healthchecks results are pushed into a Golang channel. The size of the channel is limited. This metric indicates the channel size. If healthchecks are not pushed fast enough to the exporters, this channel could be full.healthchecks
healthcheck_duration_seconds_bucket: The duration for the healthchecks executions. Labels are status, which indicates if the healthchecks were successful or not, and name which is the healthcheck name.healthcheck_duration_seconds_count: The number of execution for each healthchecks. Labels are the same than the previous metric.API
http_requests_duration_second_bucket: The duration of HTTP requests. Labels are method, which is the HTTP method of the request, and path which is the HTTP path of the request.http_requests_duration_second_count: The number of HTTP requests. Labels are the same than the previous metric.http_responses_total: A counter for HTTP responses Labels are method, path and also status, which is the response HTTP status.Go metrics
Metrics prefixed by go are golang internal metrics.
If you enable the Kubernetes service discovery module, the metrics of the Kubernetes controllers will be available on /discovery/kubernetes/metrics.
Cabourotte provides the /health and /healthz endpoints. You can use them to check if Cabourotte is started or not.