Overview of Database Connection

After deploying KubeBlocks and creating clusters, the databases run on Kubernetes, with each replica running in a Pod and managed by the InstanceSet. You can connect to the database using client tools or SDKs through the exposed database Service addresses (ClusterIP, LoadBalancer, or NodePort). See Connect database in a production environment.

If you’ve created a database using KubeBlocks in a playground or test environment, you can also use kubectl port-forward to map the database service address to a local port on your machine. Then, you can connect to the database using a client tool or the common database clients integrated inside kbcli. However, please note that this is a temporary way to access services within the cluster and is intended for testing and debugging purposes only; it should not be used in a production environment. See Connect database in a testing environment.