For more than two years running PostgreSQL with Patroni on AWS has served us very well: Scaling with the number of teams, full control over our PostgreSQL deployment and flexibility in the hardware selection including AWS i3 instances made it successful.
With the end of 2016 we started to target our next generation infrastructure running on Kubernetes, diving into how to efficiently operate PostgreSQL on Kubernetes and increase user happiness by better integrating into the Zalando tech platform and providing PostgreSQL as a Service. This talk will take you briefly through the past and current setup and will then dive deeper into the topic of our future PostgreSQL platform explaining how we plan to run PostgreSQL and which steps we take to integrate it and last but not least share with you the concept and ideas behind our open-source operator that already today drives all non-production databases and the first production databases on Kubernetes.
Jan joined the database engineering team at Zalando more than 6 years ago and learned first hand what it means to operate PostgreSQL in a data center at a larger scale in a highly available environment. Turning his back on databases he worked in various software projects including Zalando's ZMON and migrations to AWS for the last 2.5 years before joining the database engineering team in January of 2017 again. This time to work with the team to build the future databases as a service infrastructure on Kubernetes.