How to Set Up Kubernetes Monitoring with Thanos

Setting up Kubernetes monitoring with Thanos involves deploying Thanos components alongside Prometheus to achieve high availability and long-term storage of metrics. Here’s a step-by-step guide:

Prerequisites:

1. A Kubernetes Cluster: Ensure you have a running Kubernetes cluster.
2. kubectl: Make sure you have kubectl installed and configured to interact with your cluster.
3. Helm: Install Helm, a package manager for Kubernetes, if you haven’t already.

Step 1: Install Prometheus Operator

helm repo add stable https://charts.helm.sh/stable
helm repo update
helm install prometheus-operator stable/prometheus-operator

Step 2: Deploy Thanos Sidecar

Create a file named thanos-sidecar.yaml with the following content:

apiVersion: apps/v1
kind: Deployment
metadata:
  name: thanos-sidecar
spec:
  selector:
    matchLabels:
      app: thanos-sidecar
  template:
    metadata:
      labels:
        app: thanos-sidecar
    spec:
      containers:
      - name: thanos-sidecar
        image: quay.io/thanos/thanos:v0.21.0
        args:
        - "sidecar"
        - "--log.level=info"
        - "--http-address=0.0.0.0:19191"
        - "--grpc-address=0.0.0.0:10901"
        - "--reloader.config-file=/etc/reloader/reload.yaml"
        ports:
        - containerPort: 19191
        - containerPort: 10901
        volumeMounts:
        - name: config-volume
          mountPath: /etc/reloader
      volumes:
      - name: config-volume
        configMap:
          name: thanos-reloader

Apply the deployment:

kubectl apply -f thanos-sidecar.yaml

Step 3: Create Reloader ConfigMap

Create a file named thanos-reloader.yaml:

apiVersion: v1
kind: ConfigMap
metadata:
  name: thanos-reloader
data:
  reload.yaml: |
    rules:
    - source_labels: [__address__]
      action: keep
      regex: prometheus-thanos/prometheus

Apply the ConfigMap:

kubectl apply -f thanos-reloader.yaml

Step 4: Deploy Thanos Receive Component

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm install thanos-receive prometheus-community/prometheus --set-file "alertmanager.thanos-receive.yaml=thanos-receive.yaml"

Step 5: Deploy Thanos Store Component

helm install thanos-store prometheus-community/prometheus --set-file "alertmanager.thanos-store.yaml=thanos-store.yaml"

Step 6: Create Ingress for Thanos UI (Optional)

If you want to access Thanos UI from outside the cluster, create an Ingress resource.

Step 7: Access Thanos UI

Use kubectl get svc thanos-store-prometheus-community-thanos to get the NodePort of Thanos UI. Access it via http://:.

Tips:

• This is a basic setup. Depending on your needs, you might want to customize configurations for Thanos components.
• For production use, ensure you have proper authentication, authorization, and security configurations in place.
• Always refer to the official Thanos documentation for best practices and advanced configurations.

Please note that this is a simplified guide and actual configurations may vary based on your specific environment and requirements. Always refer to the official documentation for Thanos and Prometheus Operator for best practices and advanced configurations.