Different Approaches of building Stateful Cloud-Native Applications

During the early days of Kubernetes, the focus was on it as a platform to run web-scale, stateless services. Today, in just a short span of less than 7 years, we witness the explosive growth of Kubernetes as it has become the foundation of modern infrastructure for cloud computing. Stateful services such as database and analytics workloads used to run on virtual machines or cloud-based managed services. As Kubernetes is becoming ubiquitous, we are also seeing that the ecosystem is starting to address the needs of stateful applications. On the programming level, while the 12-factor methodologies for Cloud-Native applications seem to be all about stateless containerized microservices, the Saga pattern can be utilized to help with maintaining data consistency and managing transactions. We will take a look at the different approaches and key building blocks in Kubernetes storage and controllers, as well as in the Cloud-Native application layer, to understand how we can preserve the state of data as it travels across multiple network boundaries in a highly distributed computing environment.