MetricsOps algebra capable of recording Prometheus metrics
For example, the following code would wrap a org.http4s.HttpRoutes with a org.http4s.server.middleware.Metrics
that records metrics to a given metric registry.
import org.http4s.client.middleware.Metrics
import org.http4s.client.prometheus.Prometheus
val meteredRoutes = Metrics[IO](Prometheus(registry, "server"))(testRoutes)
Analogously, the following code would wrap a org.http4s.client.Client with a org.http4s.client.middleware.Metrics
that records metrics to a given metric registry, classifying the metrics by HTTP method.
import org.http4s.client.middleware.Metrics
import org.http4s.client.prometheus.Prometheus
val classifierFunc = (r: Request[IO]) =>Some(r.method.toString.toLowerCase)
val meteredClient = Metrics(Prometheus(registry, "client"), classifierFunc)(client)
Creates a MetricsOps that supports Prometheus metrics
*
* @param registry a metrics collector registry
* @param prefix a prefix that will be added to all metrics
MetricsOps algebra capable of recording Prometheus metrics
For example, the following code would wrap a org.http4s.HttpRoutes with a org.http4s.server.middleware.Metrics that records metrics to a given metric registry.
Analogously, the following code would wrap a org.http4s.client.Client with a org.http4s.client.middleware.Metrics that records metrics to a given metric registry, classifying the metrics by HTTP method.
Registers the following metrics:
{prefix}_response_duration_seconds{labels=classifier,method,phase} - Histogram
{prefix}_active_request_count{labels=classifier} - Gauge
{prefix}_request_count{labels=classifier,method,status} - Counter
{prefix}_abnormal_terminations{labels=classifier,termination_type} - Histogram
Labels --
method: Enumeration values: get, put, post, head, move, options, trace, connect, delete, other
phase: Enumeration values: headers, body
code: Enumeration values: 1xx, 2xx, 3xx, 4xx, 5xx
termination_type: Enumeration values: abnormal, error, timeout