Kubernetes monitoring service discovery¶
Kubernetes monitoring service discovery enables monitoring systems like Prometheus to dynamically detect and track targets within a cluster.^[400-devops-03-containerization-prometheus.md]
In containerized environments, IP addresses are ephemeral, making static configurations impractical. Discovery mechanisms allow the monitoring server to automatically find new services and pods, ensuring that Metrics collection continues uninterrupted as applications scale or restart.^[400-devops-03-containerization-prometheus.md]
Implementation¶
While service discovery is a core function of Prometheus in Kubernetes environments, specific configuration details often depend on the underlying deployment method.^[400-devops-03-containerization-prometheus.md] The Prometheus Operator is a common tool used to manage these configurations.^[400-devops-03-containerization-prometheus.md]
For standard Helm chart deployments, the target for scraping Metrics is typically exposed via a Kubernetes Service internal DNS name.^[400-devops-03-containerization-prometheus.md]
Configuration¶
When connecting Prometheus to data sources (such as Grafana querying Prometheus), the service discovery endpoint is referenced using the cluster's internal DNS resolution.^[400-devops-03-containerization-prometheus.md]
Example connection strings:
* Service DNS: http://prometheus-server:80
* Cluster IP: http://10.111.94.209:80^[400-devops-03-containerization-prometheus.md]
Related Concepts¶
Sources¶
^[400-devops-03-containerization-prometheus.md]