Introduction

What is an instance template

An instance serves as the fundamental unit in KubeBlocks, comprising a Pod along with several auxiliary objects. To simplify, you can initially think of it as a Pod, and henceforth, we'll consistently refer to it as an "Instance."

Starting from version 0.9, we're able to establish multiple instance templates for a particular component within a cluster. These instance templates include several fields such as Name, Replicas, Annotations, Labels, Env, Tolerations, NodeSelector, etc. These fields will ultimately override the corresponding ones in the default template (originating from ClusterDefinition and ComponentDefinition) to generate the final template for rendering the instance.

Why do we introduce the instance template

In KubeBlocks, a Cluster is composed of several Components, where each Component ultimately oversees multiple Pods and auxiliary objects.

Prior to version 0.9, these pods were rendered from a shared PodTemplate, as defined in either ClusterDefinition or ComponentDefinition. However, this design can’t meet the following demands:

• For Clusters rendered from the same addon, setting separate scheduling configurations such as NodeName, NodeSelector, or Tolerations.
• For Components rendered from the same addon, adding custom Annotations, Labels, or ENV to the Pods they manage.
• For Pods managed by the same Component, configuring different CPU, Memory, and other Resource Requests and Limits.

With various similar requirements emerging, the Cluster API introduced the Instance Template feature from version 0.9 onwards to cater to these needs.