DataStax's Cassandra Kubernetes operator reaches GA, aims to simplify deploying clusters at scale

Cassandra is an ideal data store for Kubernetes, company says

DataStax, the data management company built on the Apache Cassandra database, has made its Cassandra Kubernetes operator generally available.

Cassandra Kubernetes operator v1.0 is a collaborative effort with contributions from DataStax developers and from companies including Sky, Orange, Netflix and Target. It aims to simplify the process of deploying and managing Cassandra or DataStax Enterprise (DSE) in a Kubernetes cluster.

Kubernetes use is growing rapidly, with a recent CNCF survey noting its presence in 78 per cent of enterprises polled, compared with 58 per cent 12 months previously. It's particularly useful for deploying cloud-native containerised applications that can run anywhere. Cassandra, as a highly scalable, distributed NoSQL database with built-in replication, is a good fit for the sorts of workloads typically deployed on Kubernetes and its architecture corresponds to the native structures in Kubernetes. But Kubernetes workloads are ephemeral and application- rather than infrastructure-centric, creating a challenge when managing a large number of Kubernetes clusters together with the Cassandra clusters that hold the data.

Operators are extensions to Kubernetes that make use of the Kubernetes API or third-party custom resources to automate the management of applications. They allow administrators to configure desired behaviour in a YAML file which is then uploaded to the platform. A number of operators are published on the site OperatorHub.io, including one for Cassandra clusters by Instaclustr; others are to be found on GitHub. But DataStax is aiming to foster a community-led approach in which many developers contribute, particularly from large organisations that will be the main users, in the hope that this will become the de facto standard.

"DataStax spent the last decade solving hard data problems for distributed systems," said chief strategy officer Sam Ramji in a statement. "That has standardised and we're now looking at the decade of scale-out, cloud-native data."

The Cassandra Kubernetes operator promises "zero downtime and zero lock in". It is already deployed in the DBaaS offering DataStax Astra.

In another move that underscores the company's dual focus on epic scale and the open-source Cassandra community, DataStax recently acquired New Zealand-headquartered consultancy The Last Pickle which specialises in deploying large Cassandra clusters in companies such as AT&T, Spotify and T-Mobile.

"It's their ability to look over the horizon and handle ten times, a hundred times, a thousand times bigger environments that than the average user today," Ramji told Computing. "Big is different, you have to rearchitect. So, they had to go and write a bunch of technology to automate the solution to those problems. That automation is powerful and it's something we're excited to offer to DataStax customers and most importantly for free to the open-source community."

DataStax also announced the release of version 6.8 of its DataStax Enterprise platform.