Apache Cassandra is an open-source distributed NoSQL database management system designed to handle large amounts of data across many commodity servers. This deployment uses K8ssandra Operator, which provides a Kubernetes-native way to manage Cassandra clusters.

K8ssandra is a cloud-native distribution of Apache Cassandra that runs on Kubernetes. It includes automation for operational tasks such as repairs, backups, and monitoring.

Installation

To install Cassandra, run:

make install

Usage

After installation, you can create a Cassandra cluster:

# Check if operator is running
kubectl get pods -n cassandra

# Create a Cassandra cluster
kubectl apply -f - <<EOF
apiVersion: k8ssandra.io/v1alpha1
kind: K8ssandraCluster
metadata:
  name: demo
  namespace: cassandra
spec:
  cassandra:
    serverVersion: "4.0.1"
    datacenters:
      - metadata:
          name: dc1
        size: 3
        storageConfig:
          cassandraDataVolumeClaimSpec:
            storageClassName: standard
            accessModes: ["ReadWriteOnce"]
            resources:
              requests:
                storage: 10Gi
        config:
          jvmOptions:
            heapSize: 1Gi
EOF

Configuration

The default configuration includes:

K8ssandra Operator for managing Cassandra clusters
Support for Cassandra 4.x
Medusa for backup management
Reaper for repair scheduling
Metrics collection via Prometheus

Features

Automated Repairs: Reaper handles repair scheduling
Backup/Restore: Medusa provides backup and restore capabilities
Monitoring: Integrated Prometheus metrics
Multi-DC Support: Deploy across multiple data centers

Connecting to Cassandra

# Get CQLSH access
kubectl exec -it demo-dc1-default-sts-0 -n cassandra -c cassandra -- cqlsh

Uninstall

To uninstall:

make uninstall