README
¶
chi-prometheus (Forked*)
Prometheus middleware for chi.
This is a port of negroni-prometheus middleware (written by Rene Zbinden).
Why
Instead of logging request times, it is considered best practice to provide an endpoint for instrumentation tools (like prometheus).
How to Install our package
go get github.com/lggomez/chi-prometheus
Usage
Take a look at the example.
What do you get
An endpoint with the following information (stripped output):
# HELP chi_request_duration_milliseconds How long it took to process the request, partitioned by status code, method and HTTP path.
# TYPE chi_request_duration_milliseconds histogram
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/metrics",service="serviceName",le="300"} 1
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/metrics",service="serviceName",le="1200"} 1
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/metrics",service="serviceName",le="5000"} 1
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/metrics",service="serviceName",le="+Inf"} 1
chi_request_duration_milliseconds_sum{code="OK",method="GET",path="/metrics",service="serviceName"} 2.003123
chi_request_duration_milliseconds_count{code="OK",method="GET",path="/metrics",service="serviceName"} 1
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/ok",service="serviceName",le="300"} 0
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/ok",service="serviceName",le="1200"} 0
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/ok",service="serviceName",le="5000"} 2
chi_request_duration_milliseconds_bucke{code="OK",method="GET",path="/ok",service="serviceName",le="+Inf"} 2
chi_request_duration_milliseconds_sum{code="OK",method="GET",path="/ok",service="serviceName"} 4747.529026
chi_request_duration_milliseconds_count{code="OK",method="GET",path="/ok",service="serviceName"} 2
# HELP chi_requests_total How many HTTP requests processed, partitioned by status code, method and HTTP path.
# TYPE chi_requests_total counter
chi_requests_total{code="OK",method="GET",path="/metrics",service="serviceName"} 1
chi_requests_total{code="OK",method="GET",path="/ok",service="serviceName"} 2
Pattern Middleware Usage
Take a look at the example.
What do you get
An endpoint with the following information (stripped output):
# HELP chi_pattern_request_duration_milliseconds How long it took to process the request, partitioned by status code, method and HTTP path (with patterns).
# TYPE chi_pattern_request_duration_milliseconds histogram
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/metrics",service="test_service",le="300"} 1
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/metrics",service="test_service",le="1200"} 1
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/metrics",service="test_service",le="5000"} 1
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/metrics",service="test_service",le="+Inf"} 1
chi_pattern_request_duration_milliseconds_sum{code="OK",method="GET",path="/metrics",service="test_service"} 0.975926
chi_pattern_request_duration_milliseconds_count{code="OK",method="GET",path="/metrics",service="test_service"} 1
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/users/{firstName}",service="test_service",le="300"} 0
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/users/{firstName}",service="test_service",le="1200"} 0
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/users/{firstName}",service="test_service",le="5000"} 2
chi_pattern_request_duration_milliseconds_bucket{code="OK",method="GET",path="/users/{firstName}",service="test_service",le="+Inf"} 2
chi_pattern_request_duration_milliseconds_sum{code="OK",method="GET",path="/users/{firstName}",service="test_service"} 4755.052755
chi_pattern_request_duration_milliseconds_count{code="OK",method="GET",path="/users/{firstName}",service="test_service"} 2
# HELP chi_pattern_requests_total How many HTTP requests processed, partitioned by status code, method and HTTP path (with patterns).
# TYPE chi_pattern_requests_total counter
chi_pattern_requests_total{code="OK",method="GET",path="/metrics",service="test_service"} 1
chi_pattern_requests_total{code="OK",method="GET",path="/users/{firstName}",service="test_service"} 2
Fork Information !!!
This project was forked from edjumacator/chi-prometheus that had been forked from 766b with the intention of adding pattern matching to the logger. Original Project Here. This fork adds the possibility to register chi middleware handler to non-default prometheus registerer.
Documentation
¶
Index ¶
- func NewMiddleware(name string, buckets ...float64) func(next http.Handler) http.Handler
- func NewPatternMiddleware(name string, buckets ...float64) func(next http.Handler) http.Handler
- func NewRegMiddleware(r prometheus.Registerer, name string, buckets ...float64) func(next http.Handler) http.Handler
- func NewRegPatternMiddleware(r prometheus.Registerer, name string, buckets ...float64) func(next http.Handler) http.Handler
- type Middleware
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
func NewMiddleware ¶
NewMiddleware returns a new prometheus Middleware handler. Prometheus counters are registered to the default prometheus registerer (prometheus.DefaultRegisterer)
func NewPatternMiddleware ¶
NewPatternMiddleware returns a new prometheus Middleware handler that groups requests by the chi routing pattern. Ex: /users/{firstName} instead of /users/bob Prometheus counters are registered to the default prometheus registerer (prometheus.DefaultRegisterer)
func NewRegMiddleware ¶
func NewRegMiddleware(r prometheus.Registerer, name string, buckets ...float64) func(next http.Handler) http.Handler
NewRegMiddleware returns a new prometheus Middleware handler.
func NewRegPatternMiddleware ¶
func NewRegPatternMiddleware(r prometheus.Registerer, name string, buckets ...float64) func(next http.Handler) http.Handler
NewRegPatternMiddleware returns a new prometheus Middleware handler that groups requests by the chi routing pattern. Ex: /users/{firstName} instead of /users/bob
Types ¶
type Middleware ¶
type Middleware struct {
// contains filtered or unexported fields
}
Middleware is a handler that exposes prometheus metrics for the number of requests, the latency and the response size, partitioned by status code, method and HTTP path or path pattern
Source Files
Directories