Cost Saving - Tag - Lorenzo's Bloghttps://www.k8s.it/tags/cost-saving/Cost Saving - Tag - Lorenzo's BlogHugo -- gohugo.ioenTue, 14 Feb 2023 00:00:00 +0000HPA vs Rate-limithttps://www.k8s.it/posts/hpa-vs-rate-limit/Tue, 14 Feb 2023 00:00:00 +0000Lorenzo Girardihttps://www.k8s.it/posts/hpa-vs-rate-limit/

INTRO

Strange… we are using HPA to increase availability and introducing rate limiting to reduce it?

Well, let’s create the context.

This analysis is based on specific assumptions:

  • Cloud environment
  • Dynamic infrastructure
  • Minimum resources available

HPA

In Kubernetes, a HorizontalPodAutoscaler automatically updates a workload resource (Deployment, StatefulSet) to match demand.

Patterns

Type Behaviour
Slow and temporary Daily fluctuations, peaking during the day and troughing at night
Rapid and temporary Short bursts from poorly-behaved downstream services
Slow and persistent Request volume slowly increases as the product sees adoption
Rapid and persistent Abrupt shift from low to high volumes — e.g. called by batch jobs

Ideal Practice

Type Ideal Practice
Slow and temporary HPA should add and remove pods as necessary
Rapid and temporary HPA should NOT modify pod count — leave headroom for brief spikes
Slow and persistent HPA should add and remove pods as necessary
Rapid and persistent Leave headroom; HPA adds pods quickly to restore target utilization

Rate Limit

A rate limit is the number of API calls an app or user can make within a given time period. If this limit is exceeded — or if CPU or time limits are exceeded — the app may be throttled. Throttled requests fail.

]]>