This is grafana's answer to prometheus long term storage. By using tenants, you can isolate and lifecycle all the metrics being saved from a prometheus server into mimir, so we can save only a few days of metrics from dev while saving a lot more in production.
This a fork of grafana play with mimir along with minio's version
- clone repo:
git clone [email protected]:wick02/mimir.git && cd mimir - start up:
docker compose up
- Grafana: http://localhost:9000
- Mimir: http://localhost:9009 through 9016 for services (look in docker-compose.yml to see which routes to what)
- Prometheus: http://localhost:9091
- Ctrl C to stop the compose
docker images -q | xargs docker rmidocker ps -qa | xargs docker rmdocker volume prune