New to KubeDB? Please start here.

Horizontal Scaling Weaviate

This guide will give you an overview of how KubeDB Ops Manager scales the number of nodes (replicas) of a Weaviate cluster.

Before You Begin

• You should be familiar with the following KubeDB concepts:
  • Weaviate
  • Weaviate Quickstart

How Horizontal Scaling Process Works

The horizontal scaling process consists of the following steps:

1. At first, a user creates a Weaviate CR.

2. KubeDB provisioner operator watches for the Weaviate CR and creates a PetSet and related necessary resources.

3. Then, in order to scale the number of nodes of the Weaviate cluster, the user creates a WeaviateOpsRequest CR with the desired node count.

4. KubeDB Ops Manager watches for the WeaviateOpsRequest CR.

5. When it finds one, it halts the Weaviate object so that the KubeDB provisioner operator doesn’t perform any operation on the Weaviate during the scaling process.

6. To scale up, the Ops Manager increases the replica count of the PetSet, waits for the new nodes to join the cluster and sync the schema, and then rebalances the shard replicas across the enlarged cluster.

7. To scale down, the Ops Manager rebalances shard replicas off the nodes that are going away, then decreases the replica count of the PetSet and removes the surplus nodes.

8. After successfully scaling, the KubeDB Ops Manager updates the Weaviate object’s replica count to reflect the updated state and resumes the Weaviate object so that the KubeDB Provisioner operator resumes its usual operations.

In the next doc, we are going to show a step-by-step guide on scaling a Weaviate cluster horizontally.