This week we opened the doors on
K8ssandra, a project to run Apache Cassandra on Kubernetes in a production manner. Over time there have been multiple open source projects to facilitate this goal. From operators and metrics exporters to repair and backup interfaces. Some of this tooling has already been integrated into the Kubernetes platform, while others exist within the bounds of a Cassandra container. No matter where these hooks exist there has been a gap around pulling these pieces together into a cohesive experience. Operators have been left to glue together the pieces on their own and while many have seen success, there could be so many more.
K8ssandra takes the knowledge of how these pieces fit together and combines them into a simple collection of
Helm charts. Installation of an
Apache Cassandra cluster and complimentary tooling should be
effortless. The goal here is simplicity and to get there we will experience a number of challenges. Some will be easy, others will take more time. This is compounded by the rapidly evolving cloud-native ecosystem. We know that Cassandra is
the way to safe, scalable, data on Kubernetes. Let's knock out the hurdles to bootstrapping and start on the journey to realizing these goals.
We know not all of the choices made for this first version are set in stone as the best solution or will work in every environment. But where we are starting here, today, is defining the operator experience and laying the groundwork for the future. Check out the
roadmap for an idea of where we want to go. Help is appreciated with tooling development and integration, helm charts, testing, and documentation. K8ssandra is
not just a bunch of code and docs but a movement and we want you to join us for the ride.
~Chris