myVertica  

Rebalancing Your Cluster

After you add or remove one or more nodes in your Vertica cluster, you’ll want to make sure your data is balanced across the resulting nodes. By running the Vertica rebalance process, your data storage and workload will be balanced across all nodes in the cluster.

Note: Rebalancing can be CPU-, disk-, and network-intensive because it requires a large amount of data movement.

When to Add or Remove Nodes

You may need to add one or more nodes to your cluster when:
  • You significantly increase your data size.
  • You significantly increase your workload.
  • You need to increase the K-safety in your cluster to ensure high availability.
  • You need to swap a node out of the cluster for maintenance, upgrading, or replacement.

Removing a node is less common than adding a node. You might remove a node if the cluster is over-provisioned or if you need to divert the hardware for another purpose.

Note: Vertica doesn’t let you remove a node if doing so violates the system K-safety.

Before You Add or Remove a Node

For more information about adding or removing nodes, see the Vertica documentation.

Start the Rebalance Process

You can start the rebalancing using admintools, Management Console, or with SQL functions. For more information, see Rebalancing Data Across Nodes.

Monitoring Rebalance

You can monitor the rebalance process by reviewing the following system tables:

REBALANCE_TABLE_STATUS

REBALANCE_PROJECTION_STATUS

Learn More

To learn more about rebalancing your cluster after adding or removing a node, see the following resources:

Rebalancing Data Across Nodes
Best Practices for Preparing Your Cluster for Rebalance
Understanding Rebalancing, Part 1: What Happens During Rebalancing?
Understanding Rebalancing, Part 2: Optimizing for Rebalancing