Cassandra on Kubernetes

This guide covers deploying Apache Cassandra on Kubernetes using StatefulSets and operators.

Deployment Options

Option	Complexity	Features	Best For
StatefulSet	Medium	Basic orchestration	Simple deployments
K8ssandra	Low	Full-featured operator	Production
Cass-operator	Medium	DataStax operator	Enterprise
Custom Helm	Medium	Flexible	Custom requirements

StatefulSet Deployment

Basic StatefulSet

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: cassandra
  labels:
    app: cassandra
spec:
  serviceName: cassandra
  replicas: 3
  selector:
    matchLabels:
      app: cassandra
  template:
    metadata:
      labels:
        app: cassandra
    spec:
      terminationGracePeriodSeconds: 1800
      containers:
      - name: cassandra
        image: cassandra:4.1
        ports:
        - containerPort: 7000
          name: intra-node
        - containerPort: 7001
          name: tls-intra-node
        - containerPort: 7199
          name: jmx
        - containerPort: 9042
          name: cql
        resources:
          limits:
            cpu: "2"
            memory: 4Gi
          requests:
            cpu: "1"
            memory: 2Gi
        securityContext:
          capabilities:
            add:
              - IPC_LOCK
        lifecycle:
          preStop:
            exec:
              command:
              - /bin/sh
              - -c
              - nodetool drain
        env:
          - name: MAX_HEAP_SIZE
            value: 1G
          - name: HEAP_NEWSIZE
            value: 256M
          - name: CASSANDRA_SEEDS
            value: "cassandra-0.cassandra.default.svc.cluster.local"
          - name: CASSANDRA_CLUSTER_NAME
            value: "K8sCluster"
          - name: CASSANDRA_DC
            value: "DC1"
          - name: CASSANDRA_RACK
            value: "Rack1"
          - name: POD_IP
            valueFrom:
              fieldRef:
                fieldPath: status.podIP
        readinessProbe:
          exec:
            command:
            - /bin/bash
            - -c
            - /ready-probe.sh
          initialDelaySeconds: 15
          timeoutSeconds: 5
        volumeMounts:
        - name: cassandra-data
          mountPath: /var/lib/cassandra
  volumeClaimTemplates:
  - metadata:
      name: cassandra-data
    spec:
      accessModes: ["ReadWriteOnce"]
      storageClassName: fast
      resources:
        requests:
          storage: 100Gi

Headless Service

apiVersion: v1
kind: Service
metadata:
  name: cassandra
  labels:
    app: cassandra
spec:
  ports:
  - port: 9042
    name: cql
  clusterIP: None
  selector:
    app: cassandra

Client Service

apiVersion: v1
kind: Service
metadata:
  name: cassandra-client
  labels:
    app: cassandra
spec:
  ports:
  - port: 9042
    name: cql
  selector:
    app: cassandra

K8ssandra Operator

Installation

# Add Helm repository
helm repo add k8ssandra https://helm.k8ssandra.io/stable
helm repo update

# Install operator
helm install k8ssandra-operator k8ssandra/k8ssandra-operator -n k8ssandra-operator --create-namespace

AxonOps K8ssandra Integration

AxonOps provides production-ready container images that combine Apache Cassandra with K8ssandra Management API and AxonOps monitoring, optimized for Kubernetes deployments.

Components

Component	Description
Apache Cassandra 5.0.x	Database engine (versions 5.0.1 – 5.0.7)
K8ssandra Management API	Operational control interface
AxonOps Agent	Monitoring and management integration
cqlai	Modern CQL shell
jemalloc	Optimized memory allocator

Image Versioning

Images use a three-component versioning scheme for full immutability:

{CASSANDRA}-v{K8SSANDRA_API}-{AXONOPS}

Example: 5.0.7-v0.1.110-1.0.0

Production Deployments

Pin to specific immutable versions rather than floating latest tags. Digest-based references provide cryptographic guarantees against supply chain attacks.

Required Configuration

Variable	Description
AXON_AGENT_KEY	AxonOps API authentication key
AXON_AGENT_ORG	Organization identifier
AXON_AGENT_HOST	Server endpoint (default: agents.axonops.cloud)

Quick Start

1. Install K8ssandra Operator using the provided script
2. Configure AxonOps credentials as environment variables
3. Deploy using the example cluster configuration

Documentation and Examples

For detailed deployment instructions, configuration options, and production best practices:

Repository: github.com/axonops/axonops-containers/tree/development/k8ssandra

The repository includes:

• Installation scripts for K8ssandra Operator
• Example cluster configurations
• CI/CD workflows for building custom images
• Security scanning and verification procedures

K8ssandraCluster Resource

apiVersion: k8ssandra.io/v1alpha1
kind: K8ssandraCluster
metadata:
  name: production-cluster
spec:
  cassandra:
    serverVersion: "4.1.3"
    datacenters:
      - metadata:
          name: dc1
        size: 3
        storageConfig:
          cassandraDataVolumeClaimSpec:
            storageClassName: fast
            accessModes:
              - ReadWriteOnce
            resources:
              requests:
                storage: 500Gi
        config:
          jvmOptions:
            heapSize: 8G
        resources:
          requests:
            cpu: 2
            memory: 16Gi
          limits:
            cpu: 4
            memory: 16Gi
  stargate:
    size: 2
  reaper:
    autoScheduling:
      enabled: true
  medusa:
    storageProperties:
      storageProvider: s3
      bucketName: cassandra-backups

Storage Configuration

Storage Classes

# AWS EBS
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: fast
provisioner: ebs.csi.aws.com
parameters:
  type: gp3
  iops: "3000"
  throughput: "125"
volumeBindingMode: WaitForFirstConsumer
allowVolumeExpansion: true

---
# GCP
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: fast
provisioner: pd.csi.storage.gke.io
parameters:
  type: pd-ssd
volumeBindingMode: WaitForFirstConsumer
allowVolumeExpansion: true

---
# Azure
apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: fast
provisioner: disk.csi.azure.com
parameters:
  skuName: Premium_LRS
volumeBindingMode: WaitForFirstConsumer
allowVolumeExpansion: true

Pod Disruption Budget

apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: cassandra-pdb
spec:
  maxUnavailable: 1
  selector:
    matchLabels:
      app: cassandra

Anti-Affinity Rules

spec:
  template:
    spec:
      affinity:
        podAntiAffinity:
          requiredDuringSchedulingIgnoredDuringExecution:
          - labelSelector:
              matchExpressions:
              - key: app
                operator: In
                values:
                - cassandra
            topologyKey: kubernetes.io/hostname
          preferredDuringSchedulingIgnoredDuringExecution:
          - weight: 100
            podAffinityTerm:
              labelSelector:
                matchExpressions:
                - key: app
                  operator: In
                  values:
                  - cassandra
              topologyKey: topology.kubernetes.io/zone

Scaling

Scale Up

# Scale StatefulSet
kubectl scale statefulset cassandra --replicas=5

# Or edit the resource
kubectl edit statefulset cassandra

Scale Down

# Decommission node first
kubectl exec cassandra-4 -- nodetool decommission

# Wait for streaming to complete
kubectl exec cassandra-4 -- nodetool netstats

# Then scale down
kubectl scale statefulset cassandra --replicas=4

Monitoring

Prometheus Integration

apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
  name: cassandra
spec:
  selector:
    matchLabels:
      app: cassandra
  endpoints:
  - port: metrics
    interval: 15s

Best Practices

Resource Management

resources:
  requests:
    cpu: "2"
    memory: "8Gi"
  limits:
    cpu: "4"
    memory: "8Gi"  # Same as request for predictable performance

Health Checks

readinessProbe:
  exec:
    command:
    - /bin/bash
    - -c
    - "nodetool status | grep -E '^UN\\s+${POD_IP}'"
  initialDelaySeconds: 90
  periodSeconds: 30
  timeoutSeconds: 10

livenessProbe:
  exec:
    command:
    - /bin/bash
    - -c
    - "nodetool info"
  initialDelaySeconds: 120
  periodSeconds: 30
  timeoutSeconds: 10

---

Next Steps

• AWS Deployment - EKS specifics
• GCP Deployment - GKE specifics
• Operations - Kubernetes operations