autoscale

Scaling Kubernetes Resources

Creates an autoscaler that automatically chooses and sets the number of pods that run in a kubernetes cluster.

Looks up a Deployment, ReplicaSet, StatefulSet, or ReplicationController by name and creates an autoscaler that uses the given resource as a reference. An autoscaler can automatically increase or decrease number of pods deployed within the system as needed.

Command

$ kubectl autoscale (-f FILENAME | TYPE NAME | TYPE/NAME) [--min=MINPODS] --max=MAXPODS [--cpu-percent=CPU]

[OR]

$ kubectl hpa (-f FILENAME | TYPE NAME | TYPE/NAME) [--min=MINPODS] --max=MAXPODS [--cpu-percent=CPU]

hpa stands for Horizontal Pod Autoscale

Example

Current State

$ kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
nginx-6db489d4b7-zcc8h   1/1     Running   0          5s

Command

$ kubectl autoscale deployment nginx --min=2 --max=10 --cpu-percent=80
horizontalpodautoscaler.autoscaling/nginx autoscaled

[OR]

$ kubectl hpa deployment nginx --min=2 --max=10 --cpu-percent=80
horizontalpodautoscaler.autoscaling/nginx autoscaled

This will make sure to auto-scale horizontally when the CPU usage hits 80%.

Output

$ kubectl get pods
NAME                     READY   STATUS    RESTARTS   AGE
nginx-6db489d4b7-2rrrm   1/1     Running   0          15s
nginx-6db489d4b7-vxqwm   1/1     Running   0          53s

Notice that the command has an arg that says --min=2, the deployment instantaniously auto-scales to 2 pods.