History

* move out prometheus metrics
* not use gorilla wrappers

Signed-off-by: Vasiliy Tolstov <v.tolstov@unistack.org>

2020-08-25 17:59:41 +03:00

noop

regen files with never protoc (#6 )

2020-08-19 17:47:17 +03:00

wrapper

regen files with never protoc (#6 )

2020-08-19 17:47:17 +03:00

options.go

Metrics interface and Prometheus implementation (#1929 )

2020-08-18 08:27:50 +01:00

README.md

Metrics interface and Prometheus implementation (#1929 )

2020-08-18 08:27:50 +01:00

reporter.go

Update reporter.go

2020-08-18 08:30:29 +01:00

README.md

metrics

The metrics package provides a simple metrics "Reporter" interface which allows the user to submit counters, gauges and timings (along with key/value tags).

Implementations

Prometheus (pull): will be first
Prometheus (push): certainly achievable
InfluxDB: could quite easily be done
Telegraf: almost identical to the InfluxDB implementation
Micro: Could we provide metrics over Micro's server interface?

Todo

Include a handler middleware which uses the Reporter interface to generate per-request level metrics
- Throughput
- Errors
- Duration