Complete Guide to Run Cassandra on Kubernetes

Follow the step by step guide below to deploy Cassandra on Kubernetes.

Why Cassandra on Kubernetes

Cassandra as a NoSQL database is well suited to the high availability and scalability needs of cloud based applications and it is massively adopted now. With the skyrocketing popularity of Kubenetes, it is no surprise that many developers are running Cassandra on Kubernetes for managing data on Kubernetes.

Deploying Cassandra on Kubernetes

mountainmountainmountainmountainmountain
1. Prerequisites
2. Pre Config
3. Get started
4. Install Cassandra
5. Monitoring

Cassandra Operators

MayaData supports multiple operators to run Cassandra on Kubernetes

Kudo Operators

Kudo Operators

Kubernetes Universal Declarative Operator is a toolkit that helps to manage stateful applications after they are deployed on Kubernetes Cluster. While Kubernetes already comes in with a lot of built in automation to run simple workloads, KUDO is designed to help complex scenarios.

DataStax Operators

DataStax Operators

DataStax is a hybrid database-as-a-service built on Apache Cassandra. DataStax built on Apache Cassandra gives you the freedom to run your data on any device and in any cloud.

OpenEBS Benefits for Cassandra

If you are running stateful applications like Cassandra, then OpenEBS and its various storage engines can help in simplifying the usage in many ways:

Persistent Volumes

Persistent Volumes

OpenEBS Dynamic LocalPV provides persistent volumes on the fly. With OpenEBS persistent volumes provisioned dynamically, disks and cloud volumes can be discovered, pooled, and provisioned.

Low Latency Performance

Low Latency Performance

Recent benchmarking with Datastax Cassandra & OpenEBS showed extreme low latency performance in the range of 1.5-2.0 ms and high row & partition rates of upto 3 Million writes on Datastax Cassandra.

Highly available storage

Highly available storage

In some cases such as Cassandra deployments where the replication factor is low or where the cost of recovering and bootstrapping a node may be high, it may be beneficial to use block level synchronous replicas from OpenEBS.

Increased density

Increased density

Through the use of OpenEBS localPV and Kubernetes, Cassandra can be safely provisioned in such a way that nodes are shared. In this way for example multiple Cassandra rings can share the same underlying devices while OpenEBS addresses the scheduling to ensure data locality, provisioning, device pooling and more. This increased density can dramatically improve the economics and decrease the carbon footprint of a deployment.

Useful Cassandra Resources

Solution Guide

DataStax Cassandra using OpenEBS LVM LocalPV on EKS

In this guide we will deploy Datastax Cassandra on OpenEBS LVM LocalPV. DataStax Kubernetes Operator for Apache Cassandra® (Cass Operator) automates the process of deploying and managing open-source Apache Cassandra® or DataStax Enterprise (DSE) in a Kubernetes cluster.

Solution Guide

Datastax Cassandra using OpenEBS ZFS Local PV in EKS

All you need is a Kubernetes Cluster. We will be using EKS where we will install Datastax Cassandra using the Kubectl method.

Solution Guide

Datastax Cassandra using OpenEBS ZFS Local PV in GKE

We will be using GKE where we will install Datastax Cassandra using the Kubectl method.

Docs

OpenEBS for Cassandra

In this docs page, we will talk about the advantages of using OpenEBS for Cassandra database, deployment model and configuration workflow.

Blog

How to deploy a Cassandra Cluster on Kubernetes + OpenEBS

In this blog, we will learn about the steps to deploy a cassandra cluster on Kubernetes with the help of OpenEBS.

Docs

Cassandra Pod delete experiment details

Gain confidence in how your app reacts when a container crashes or is forcefully shutdown.

Cassandra Events & Webinar

MayaData regularly attends worldwide events and organizes resourceful webinars to interact with our audience to share knowledge & resources and announce exciting product updates, launches and releases.

Past Events

Operating Cassandra on Kubernetes with the help of OpenEBS & LitmusChaos September 24, 2020

Webinar: Operating Cassandra on Kubernetes with the help of OpenEBS & LitmusChaos

With Kubernetes popularity skyrocketing and the massive adoption of Apache Cassandra as a NoSQL database well-suited to the high availability and scalability needs of cloud-based applications, it should be no surprise that more developers are running Cassandra on Kubernetes. However, many devs are searching for the right solutions to use to ease the operations of Cassandra on Kubernetes.

Data Protection for Kubernetes August 6, 2020

Webinar: Data Protection for Kubernetes

Kubernetes and the use of Container Attached Storage such as OpenEBS enables per workload management of data. This approach has quickly become a standard way to preserve and protect the benefits of Kubernetes for data. Small, loosely coupled teams manage small, loosely coupled workloads. There can be 150 or more such workloads per a cluster, and each of these workloads and teams may have different back-up and disaster recovery requirements.

Kubernetes and storage. Kubernetes for storage. An overview. July 16, 2020

CNCF Member Webinar: Kubernetes and storage. Kubernetes for storage. An overview.

Originally designed for stateless workloads only, over the last two to three years, the growth of stateful workloads on Kubernetes has been rapid. Today it is common to see monitoring, logging, e-commerce, CI/CD, ML, and other workloads running on Kubernetes.

Kubernetes native management of Kubernetes native data June 24, 2020

Kubera - Kubernetes native management of Kubernetes native data

Originally designed for stateless workloads only, over the last two to three years, the growth of stateful workloads on Kubernetes has been rapid. Today it is common to see monitoring, logging, e-commerce, CI/CD, ML, and other workloads running on Kubernetes.

OpenEBS - Still Free and now FASTEST Kubernetes Storage May 28, 2020

OpenEBS - Still Free and now FASTEST Kubernetes Storage

Shared storage systems have fallen out of favor in part because their performance is difficult to characterize, unstable, and generally much slower than underlying NVMe systems. The tight coupling of workloads together via shared storage and the resulting blast radius is also worthy of consideration.

Enterprise Support

Need enterprise support for OpenEBS under Cassandra and other workloads?