A streamlined solution to deploy and configure Prometheus monitoring and alerting for Kubernetes clusters across AWS, GCP, and Azure without using Helm.
Monitoring Kubernetes clusters on AWS, GCP and Azure using Prometheus Operator and Grafana
This tool is designed for Kubernetes administrators and DevOps teams who want to quickly set up comprehensive monitoring and alerting for their clusters running on major cloud providers. It simplifies deploying Prometheus Operator, Alertmanager, and Grafana with preconfigured dashboards and alerts, enabling effective network monitoring and security automation in cloud environments.
Users should commit any custom configuration changes before deployment as the deploy script may overwrite files. SMTP and Slack tokens are optional but recommended for alert notifications. Proper security group port openings are essential for metrics collection. For Kubernetes 1.12+, cAdvisor deployment is required to gather container resource usage metrics.
Ensure you have a Kubernetes cluster and kubectl configured
Configure Security Groups to allow ports 9100, 10250, 10251, 10252, 10054, and 10055
For Kubernetes 1.12+, deploy cAdvisor bound to host port 4194
Clone the repository and checkout the latest release using: curl -L https://git.io/getPrometheusKubernetes | sh -
Customize component versions and settings via the interactive deployment script or by editing manifest files
Run the deployment script: ./deploycurl -L https://git.io/getPrometheusKubernetes | sh -
Clones the repository and checks out the latest release
./deploy
Deploys the complete Prometheus monitoring stack on the Kubernetes cluster
kubectl port-forward
Access Prometheus and Grafana dashboards locally by forwarding ports
./teardown
Removes all deployed monitoring components from the cluster
scripts/generate-rules-configmap.sh
Generates updated Prometheus alert rules configmap after modifying alert rules
kubectl apply -f manifests/prometheus/prometheus-k8s-rules.yaml -n monitoring
Applies updated Prometheus alert rules to the monitoring namespace