Auto Scale
Auto-scaling is a powerful method for automatically adjusting the number of instances or resources allocated to a service based on fluctuating workload demands. Aperture's closed-loop control policies provide an ideal solution for defining auto-scaling criteria. Aperture Agents interface with infrastructure APIs to perform auto-scaling. For example, Aperture Agents can invoke Kubernetes APIs to perform auto-scaling of any scalable resource in a Kubernetes cluster.
Setup
Aperture performs auto-scaling based on Signals in an Aperture policy.
- Kubernetes: Auto-scaling any scalable resource in a Kubernetes cluster can be achieved using Aperture. To achieve this, you must first ensure that an Aperture Agent is installed on the cluster, and then configure auto-scaling policies at the Aperture Controller.