hkford / prometheus-python-golang-rust

Prometheus instrumentation examples in Python(Flask), Golang(Echo) and Rust(Axum)

Prometheus instrumentation examples in Python, Golang and Rust

Prerequisites

• Docker
• Docker Compose V2

Prometheus server scrapes applications

Three applications are running and Prometheus server scrapes them.

Run Prometheus server and applications

$ docker compose -f compose-prometheus.yaml up

Check target status

Open http://localhost:80 and check all targets are up. Click Status -> Targets and see axum-app, echo-app, flask-app and prometheus-server-itself are up.

Send some requests

Flask application is running at port 5000, Echo applications is at port 8888 and Axum application is at port 3000. Send some requests to them.

# requests to Flask application
$ curl -XPOST http://localhost:5000/movies
{"movies":["movie42"]}
$ curl -XPOST http://localhost:5000/movies
{"movies":["movie42","movie10"]}
$ curl http://localhost:5000/movies
{"movies":["movie42","movie10"]}
# requests to Echo application
$ curl -XPOST http://localhost:8888/movies
[{"ID":881}]
$ curl -XPOST http://localhost:8888/movies
[{"ID":881},{"ID":947}]
$ curl -XPOST http://localhost:8888/movies
[{"ID":881},{"ID":947},{"ID":717}]
$ curl http://localhost:8888/movies
[{"ID":881},{"ID":947},{"ID":717}]
# requests to Axum application
$ curl -XPOST http://localhost:3000/movies
[{"id":18168806418232667828}]
$ curl -XPOST http://localhost:3000/movie
[{"id":18168806418232667828},{"id":16276746138974980245}]
$ curl http://localhost:3000/movies
[{"id":18168806418232667828},{"id":16276746138974980245}]
$ curl -XDELETE http://localhost:3000/movies/1
[{"id":18168806418232667828}]

Visualize metrics using Prometheus UI

Execute PromQL like axum_movies_count and now you can see metrics graph.

Clean up

$ docker compose -f compose-prometheus.yaml down

ADOT collector scrapes applications and send metrics to Prometheus server

Three applications are running and OpenTelemetry collector scrapes them, then it remote writes to Prometheus server.

Run Prometheus server, ADOT collector and applications

$ docker compose -f compose-adot-collector.yaml up

Check target status

You can see only prometheus-server-itself in targets.

Send some requests

Flask application is running at port 5000, Echo applications is at port 8888 and Axum application is at port 3000. Send some requests to them.

# requests to Flask application
$ curl -XPOST http://localhost:5000/movies
{"movies":["movie42"]}
$ curl -XPOST http://localhost:5000/movies
{"movies":["movie42","movie10"]}
$ curl http://localhost:5000/movies
{"movies":["movie42","movie10"]}
# requests to Echo application
$ curl -XPOST http://localhost:8888/movies
[{"ID":881}]
$ curl -XPOST http://localhost:8888/movies
[{"ID":881},{"ID":947}]
$ curl -XPOST http://localhost:8888/movies
[{"ID":881},{"ID":947},{"ID":717}]
$ curl http://localhost:8888/movies
[{"ID":881},{"ID":947},{"ID":717}]
# requests to Axum application
$ curl -XPOST http://localhost:3000/movies
[{"id":18168806418232667828}]
$ curl -XPOST http://localhost:3000/movie
[{"id":18168806418232667828},{"id":16276746138974980245}]
$ curl http://localhost:3000/movies
[{"id":18168806418232667828},{"id":16276746138974980245}]
$ curl -XDELETE http://localhost:3000/movies/1
[{"id":18168806418232667828}]

Visualize metrics using Prometheus UI

Execute PromQL like axum_movies_count and now you can see metrics graph.

Clean up

$ docker compose -f compose-adot-collector.yaml down