we're releasing an official grafana/prometheus collector for Supabase (MIT license).
The service collects 197 metrics[0], along with a default Grafana dashboard[1]. The entire thing can be run locally or deployed to a cloud service like Fly.io
Many Supabase customers already use Prom/Grafana, and this provides another way for them to debug their Postgres database.
We'd love feedback if you're missing anything important metrics
Supabase Engineer here, happy to answer any questions!
We've made a Prometheus-compatible metrics endpoint[0] available to our customers for quite a while now, which is actively used by customers to monitor their projects.
Setting up a metrics scraper and a Grafana instance, all while figuring out how and where to host it can pose a lot of friction for folks which haven't had much hands-on infrastructure experience.
This aims to cut down on that friction, and allow everybody to deploy it on Docker-compatible platforms and services, such as Fly.io.
The included dashboard has significant overlap with what Supabase is using internally to monitor and debug project infrastructure.
Love this. We have been wanting to set up a Grafana dashboard for the metrics endpoint, but never found the time to set it up, this is really helpful! Also being able to run it locally is really nice.
(small note, in the readme it says `cp .env.sample .env`, looks like it should be `cp .env.example .env`)
The service collects 197 metrics[0], along with a default Grafana dashboard[1]. The entire thing can be run locally or deployed to a cloud service like Fly.io
Many Supabase customers already use Prom/Grafana, and this provides another way for them to debug their Postgres database.
We'd love feedback if you're missing anything important metrics
[0] metrics: https://github.com/supabase/supabase-grafana/blob/main/docs/...
[1] dashboard: https://github.com/supabase/supabase-grafana/blob/main/docs/...