Investigate use of Event-driven Autoscaling (KEDA)

[bengtfredh]

About

Deployments and StatefulSets are the most common way to scale workloads with KEDA.

It allows you to define the Kubernetes Deployment or StatefulSet that you want KEDA to scale based on a scale trigger. KEDA will monitor that service and based on the events that occur it will automatically scale your resource out/in accordingly.

Behind the scenes, KEDA acts to monitor the event source and feed that data to Kubernetes and the HPA (Horizontal Pod Autoscaler) to drive rapid scale of a resource. Each replica of a resource is actively pulling items from the event source. With KEDA and scaling Deployments/StatefulSet you can scale based on events while also preserving rich connection and processing semantics with the event source (e.g. in-order processing, retries, deadletter, checkpointing).

Read more https://keda.sh/

Use case

We can as start deploy KEDA to test environment and scale deployment down to 0. When application is requested by i.e. http app will scale up 1 replica and add replicas as needed.

Acceptenace criteria

1. A description of features that can benefit us.
2. Identify and describe pros and cons
3. Propose further actions

1. A description of features that can benefit us.

2. Identify and describe pros and cons

• Azure have Vertical Pod Autoscaler in preview. We may not need one or the other, but smart thing to do is add one at a time.

3. Propose further actions