Playing with Elephants ~PostgreSQL and Pgpool-II~

For a production environment, it is important to design a disaster recovery strategy. The traditional disaster recovery strategy is to make regular database backups and store them. However, large databases may take hours or several days to restore. Cloud computing has made disaster recovery more efficient and simple. Cloud computing enables the deployment of applications or databases across multiple availability zones or multiple regions to achieve high availability and disaster recovery.  Today, many companies have migrated their applications to a microservices architecture and deploy the application systems including databases on Kubernetes. However, how to implement disaster recovery for databases deployed on Kubernetes? In this post, I'd like to describe the disaster recovery strategies for PostgreSQL deployment on Kubernetes. Because database is a stateful service, it is hard to manage and scale databases on Kubernetes. PostgreSQL Operators (e.g. Crunchy PostgreSQL Opera