Skip to content

Scaling

A storage cluster scales out by adding Kubernetes workers as storage nodes, which adds capacity and performance at the same time. New workers are described in a growth ClusterDeploymentConfig that names the existing cluster, either written by hand or drafted by a discovery run. The cluster keeps serving I/O throughout, and the data is rebalanced onto the new devices afterward.

Topics

Topic Purpose
Expanding a Storage Cluster Adding storage nodes to a running cluster and following their integration.
Parallel Storage Node Addition Provisioning several workers concurrently instead of one after another.
Managing Superclusters Combining several storage clusters into one, and placing volumes in them.