Skip to content
Chaos Lab

Challenge 44 · intermediate · Horizontal Pod Autoscaler

Autoscaling too late

An API is scaled by a Horizontal Pod Autoscaler targeting 90% CPU, checking every 15 seconds, with 2 pods minimum. Each pod handles 120 requests a second and takes 60 seconds to become ready. At 20 seconds a campaign email lands and traffic climbs from 200 to 600 requests a second over a minute.

From 20 to 120 seconds, answer at least 99% of requests, averaging no more than 8 pods.

ready00:00 / 02:00Success · 10 s–p99 · 10 s–

Goal · measured 00:20–02:00

0 changes · par 2
  • Requests answered ≥ 99.0%not measured yet
  • Average pods ≤ 8.0not measured yet

This is a simulation. Its numbers come from a model of a typical system, not from measuring real servers. The same link and seed always give the same run. How the Chaos Lab works.