This is an example on how to setup RabbitMQ so Prometheus can scrape data from it. It uses a third party RabbitMQ exporter.
Since the RabbitMQ exporter needs to scrape the RabbitMQ management API to scrape data, and it defaults to localhost, it is easier to simply embed the kbudde/rabbitmq-exporter on the same pod as RabbitMQ, this way they share the same network.
With this pod running you will have the exporter scraping data, but Prometheus has not yet found the exporter and is not scraping data from it.
After you got Kubernetes service discovery up and running you just need to advertise that RabbitMQ is exposing metrics. To do that you need to define a service that:
- Exposes the exporter port
- Has a prometheus.io/scrape: "true" annotation
- Has a prometheus.io/port: "9090" annotation
And you should be able to see your RabbitMQ exporter being scraped on the Prometheus status page. Since the IP that will be scraped will be the pod endpoint it is important that the node where Prometheus is running has access to the Kubernetes overlay network (flannel, Weave, AWS, or any of the other options that Kubernetes gives to you).