This talk is structured in 2 parts, the first is an introduction to immutable infrastructures, what it is, advantages and disadvantages and architecture requirements for applications. In the second part I'll explain how we apply those principles at the storage team by Schibsted for our stateful services to provide a managed database offering for multiple customers. The use cases will be based on Cassandra, Kafka and zookeeper which we currently support in many clusters in production.

The public cloud is making a huge impact on the way enterprises host, manage, and scale their database operations. They’ve been freed from the scalability constraints of their on-premise systems. They can provision new infrastructure at the click of a button, without a lengthy hardware procurement process. They no longer need large, upfront capital investments to launch new IT projects. And they’re taking advantage of new, cloud-oriented technologies when storing and analyzing their data. Through a range of Database as a Service (DBaaS) options, public cloud vendors now make it easier than ever for organizations to migrate and maintain their databases. But while these solutions address many of the headaches involved in database management such as migration, provisioning, and administration, there are significant differences between the various DBaaS offerings on the market. In this session, we explore the core DBaaS solutions provided by the leading cloud platforms, AWS, GCP and Microsoft Azure, and compare key features such as the types of the database on offer, Infrastructure Design Challenges, DR and HA capabilities, Performance Indicators, Pricing and Database Security.