Auto-scaling & Scaling to Zero

Because Kubeletto runs Knative, it scales container instances based on active concurrent request volume.

Scaling Configurations

• Min Instances: Minimum active instances. On the free tier, this MUST be 0 (enforcing scale-to-zero).
• Max Instances: Cap on the maximum instances the horizontal scaler can spawn.
• Concurrency Target: The threshold of concurrent requests handled by one instance before scaling out.

Deploying Scaling Limits

kubeletto deploy my-service --image my-image:latest --min-instances 0 --max-instances 3

Understanding Scale-to-Zero

When a service receives no requests for 60 seconds, Knative routes scale down container instances. The next incoming request triggers a "cold start," which loads container layers and starts your application. Cold starts usually take between 2 to 8 seconds.