Running production quality database in Kubernetes can be tricky to say the least. In the early days of Kubernetes, replication controllers were used to run a single pod for a database. With the introduction of StatefulSet, it became easy to run a docker container for any database. But what about monitoring, taking periodic backups, restoring from backups or cloning from an existing database? KubeDB is a framework for writing operators for any database that support the following operational requirements: - Create a database declaratively using TPR. - Take one-off backups or period backups to various cloud stores, eg, S3, GCS, etc. - Restore from backup or clone any database. - Native integration with Prometheus for monitoring via CoreOS Prometheus Operator. - Apply deletion lock to avoid accidental deletion of database. - Keep track of deleted databases, cleanup prior snapshots with a single command. - Use cli to manage databases like kubectl for Kubernetes.
