You are looking at the documentation of a prior release. To read the documentation of the latest release, please visit here.

New to KubeDB? Please start here.

SingleStore Compute Resource Autoscaling

This guide will give an overview on how KubeDB Autoscaler operator autoscales the database compute resources i.e. cpu and memory using singlestoreautoscaler crd.

Before You Begin

How Compute Autoscaling Works

The following diagram shows how KubeDB Autoscaler operator autoscales the resources of SingleStore database components. Open the image in a new tab to see the enlarged version.

  Compute Auto Scaling process of SingleStore
Fig: Compute Auto Scaling process of SingleStore

The Auto Scaling process consists of the following steps:

  1. At first, a user creates a SingleStore Custom Resource Object (CRO).

  2. KubeDB Provisioner operator watches the SingleStore CRO.

  3. When the operator finds a SingleStore CRO, it creates required number of PetSets and related necessary stuff like secrets, services, etc.

  4. Then, in order to set up autoscaling of the various components (ie. Aggregator, Leaf, Standalone) of the SingleStore database the user creates a SingleStoreAutoscaler CRO with desired configuration.

  5. KubeDB Autoscaler operator watches the SingleStoreAutoscaler CRO.

  6. KubeDB Autoscaler operator generates recommendation using the modified version of kubernetes official recommender for different components of the database, as specified in the SingleStoreAutoscaler CRO.

  7. If the generated recommendation doesn’t match the current resources of the database, then KubeDB Autoscaler operator creates a SingleStoreOpsRequest CRO to scale the database to match the recommendation generated.

  8. KubeDB Ops-manager operator watches the SingleStoreOpsRequest CRO.

  9. Then the KubeDB Ops-manager operator will scale the database component vertically as specified on the SingleStoreOpsRequest CRO.

In the next docs, we are going to show a step by step guide on Autoscaling of various SingleStore database components using SingleStoreAutoscaler CRD.