New to KubeDB? Please start here.

PgBouncer Vertical Scaling

This guide will give an overview on how KubeDB Ops-manager operator updates the resources(for example CPU and Memory etc.) of the PgBouncer.

Before You Begin

• You should be familiar with the following KubeDB concepts:
  • PgBouncer
  • PgBouncerOpsRequest

How Vertical Scaling Process Works

The following diagram shows how KubeDB Ops-manager operator updates the resources of the PgBouncer. Open the image in a new tab to see the enlarged version.

  

The vertical scaling process consists of the following steps:

1. At first, a user creates a PgBouncer Custom Resource (CR).

2. KubeDB Provisioner operator watches the PgBouncer CR.

3. When the operator finds a PgBouncer CR, it creates PetSet and related necessary stuff like secrets, services, etc.

4. Then, in order to update the resources(for example CPU, Memory etc.) of the PgBouncer, the user creates a PgBouncerOpsRequest CR with desired information.

5. KubeDB Ops-manager operator watches the PgBouncerOpsRequest CR.

6. When it finds a PgBouncerOpsRequest CR, it pauses the PgBouncer object which is referred from the PgBouncerOpsRequest. So, the KubeDB Provisioner operator doesn’t perform any operations on the PgBouncer object during the vertical scaling process.

7. Then the KubeDB Ops-manager operator will update resources of the PetSet to reach desired state.

8. After the successful update of the resources of the PetSet’s replica, the KubeDB Ops-manager operator updates the PgBouncer object to reflect the updated state.

9. After the successful update of the PgBouncer resources, the KubeDB Ops-manager operator resumes the PgBouncer object so that the KubeDB Provisioner operator resumes its usual operations.

Vertical Scaling Modes

KubeDB actuates vertical scaling in one of two modes, selected through the spec.verticalScaling.mode field of the PgBouncerOpsRequest:

• Restart (default): The operator patches the PetSet with the new resources and restarts the Pods (one at a time, honoring the database’s failover rules) so they come back with the updated CPU and Memory. This works on every Kubernetes cluster.
• InPlace: The operator resizes the running containers in place using the Kubernetes in-place Pod resize (pods/resize subresource) — no Pod restart, so scaling happens without downtime or failover. If a Node cannot accommodate the new resources (the resize is reported Infeasible), the operator automatically falls back to the Restart behavior for that Pod.

If spec.verticalScaling.mode is omitted, it defaults to Restart.

Note: InPlace mode relies on the Kubernetes InPlacePodVerticalScaling feature gate, which is enabled by default from Kubernetes v1.33. On older clusters, or when the feature gate is disabled, use Restart mode.

In the next docs, we are going to show a step-by-step guide on updating resources of PgBouncer PgBouncerOpsRequest CRD.