Hi, thank you for your valuable comment and for sharing your experiences.

Regarding your question about HPA, Kubernetes does not wait 5 minutes to scale-up once HPA is triggered. The scale-up process starts almost immediately after the metrics indicate the need for more pods, with a typical check interval of 15 seconds. However, for scale-down (scale-in), Kubernetes waits for 5 minutes to ensure stability.

Here are the relevant parameters and their default values:

• --horizontal-pod-autoscaler-sync-period: 15 seconds (frequency of metric checks)
• --horizontal-pod-autoscaler-upscale-delay: 3 minutes (minimum time between scale-up operations)
• --horizontal-pod-autoscaler-downscale-delay: 5 minutes (time before scaling down)