Deploy Prometheus Monitoring with Prometheus Operator

Deploy Prometheus Monitoring with Prometheus Operator

This guide outlines the deployment process for a custom Prometheus monitoring setup using the Prometheus Operator.

Prerequisites

Ensure your Kubernetes cluster has:

1. Prometheus Operator
2. Kubernetes Metrics Server
3. Helm

Deployment Steps

Install Prometheus Operator

First, we need to install the Prometheus Operator:

a. Create a namespace for monitoring:

kubectl create namespace monitoring

b. Add the Prometheus Operator Helm repository:

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update

c. Install the Prometheus Operator:

helm install prometheus-operator prometheus-community/kube-prometheus-stack -n monitoring

d. Verify the installation:

kubectl get pods -n monitoring

Configure Values

Edit values.yaml to set your specific configuration:

Deployment:
  name: <project-id>
  namespace: default

PersistentVolume:
  node: jiriaf2302-control-plane
  path: /var/prom
  size: 5Gi

Prometheus:
  serviceaccount: prometheus-k8s
  namespace: monitoring

Key configurations:

• Deployment.name: Used for job naming, persistent volume path, and service monitoring selection
• Deployment.namespace: Specifies job namespace and namespace monitoring selection
• PersistentVolume.*: Configures storage for Prometheus data
• Prometheus.*: Sets Prometheus server details

Note: Only those servicemonitors with the namespace default can be monitored. To monitor additional namespaces, additional configuration is required. Refer to the Prometheus Operator documentation on customizations for details.

Install the Custom Prometheus Helm Chart

Run the following command, replacing <project-id> with your identifier:

helm install <project-id>-prom prom/ --set Deployment.name=<project-id>

Example:

ID=jlab-100g-nersc-ornl
helm install $ID-prom prom/ --set Deployment.name=$ID

Verify Deployment

Check that all components are running:

kubectl get pods -n monitoring
kubectl get pv

Access Grafana Dashboard

a. Find the Grafana service:

kubectl get svc -n monitoring

b. Set up port forwarding:

kubectl port-forward svc/prometheus-operator-grafana -n monitoring 3000:80

c. Access Grafana at http://localhost:3000 (default credentials: admin/admin)

Components Deployed

• Prometheus Server (prometheus.yaml)
• Persistent Volume for data storage (prom-pv.yaml)
• Empty directory creation job (prom-create_emptydir.yaml)

Integration with Workflows

This setup is designed to monitor services and jobs created by your workflow system.

Advanced Configuration

For further customization, refer to the Helm chart templates and values.yaml. Ensure your cluster has the necessary permissions and resources for persistent volumes and Prometheus server operation.

Troubleshooting

If you encounter issues:

1. Check pod status: kubectl get pods -n monitoring
2. View pod logs: kubectl logs <pod-name> -n monitoring
3. Ensure persistent volume is correctly bound: kubectl get pv
4. Verify Prometheus configuration: kubectl get prometheus -n monitoring -o yaml

For more help, consult the Prometheus Operator documentation.