Building a Database as a Service on Kubernetes

As more teams and companies move to the cloud, the demand for scalable storage on the cloud necessarily increases. As a company that has undertaken this journey to provide scalable storage as a service on Kubernetes, we want to outline this process to help others in the cloud native ecosystem do the same.

Throughout the presentation, we showcase lots of different CNCF/Cloud Native projects and their impact on providing value to any complete service offering. For example, we show how the concurrent deployment of operators allows us to deploy Vitess and Prometheus with ease. We explain the power of auto-provisioning Grafana dashboards, choosing valuable metrics, and configuring them for enterprise-ready deployments. Additionally, we detail how Vitess seamlessly integrates stateful transactions in an otherwise stateless environment.

On the development side, we describe the powerful features of the operator-sdk that let us define complex Vitess deployments and spin up, reconcile, and control them with intuitive configuration. By providing context on how we developed an operator for a product with many moving parts, we hope to encourage the development and upgrading of operators in the cloud native ecosystem.