# Why K8ssandra?

**URL:** <https://forum.devtalk.com/t/why-k8ssandra/12079>\
**Category:** General Dev Chat\
**Tags:** k8ssandra, kubernetes, database, cassandra, k8s, containers\
**Created:** [6 July 2021 23:34 UTC](https://forum.devtalk.com/t/why-k8ssandra/12079 "2021-07-06T23:34:18Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Exadra37](https://forum.devtalk.com/user_avatar/forum.devtalk.com/exadra37/32/683_2.png) [@Exadra37](https://forum.devtalk.com/u/Exadra37)\
**Post date:** [6 July 2021 23:34 UTC](https://forum.devtalk.com/t/why-k8ssandra/12079/1 "2021-07-06T23:34:18Z")

</div>

> **[Why K8ssandra?](https://k8ssandra.io/blog/articles/why-k8ssandra/)**
>
> Why should you run Apache Cassandra on Kubernetes? Jeff Carpenter explores the inspiration and advantages to this model with K8ssandra.

> K8ssandra Defined
> 
> So, what is K8ssandra, exactly? K8ssandra is an open source project with the mission of capturing SRE knowledge and best practices. This knowledge is distilled into a collection of Helm charts. The charts are deployable prescriptions for how to run Cassandra, along with supporting tools that ensure smooth operation of Cassandra clusters of any size.
> 
> The core of K8ssandra is [cass-operator](https://k8ssandra.io/docs/architecture/cassandra/), a [Kubernetes operator](https://kubernetes.io/docs/concepts/extend-kubernetes/operator/) which includes a [custom resource definition (CRD)](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) for Cassandra Datacenters. Cass-operator has two fundamental roles. First, it translates logical Cassandra terms like datacenters, racks, and nodes into Kubernetes resources such as labels, stateful sets, and pods. K8ssandra deploys those resources on the Kubernetes distribution of your choice. Second, it responds to Kubernetes notifications and takes corrective actions to reconcile state changes. This includes scaling the Cassandra cluster up or down based on a change to your desired number of nodes, or reacting to a pod terminated event for a Cassandra node by creating a replacement node and attaching a storage volume containing the correct data files.
