037 Project 37: Kubernetes HPA Simulator

037 Build a Kubernetes HPA Simulator

Simulate replica scaling from CPU load over time.

Full main.go

package main

import (
    "fmt"
    "math/rand"
    "time"
)

func main() {
    replicas := 2
    targetCPU := 60.0

    for t := 1; t <= 30; t++ {
        cpu := 30 + rand.Float64()*70
        if cpu > targetCPU+10 && replicas < 20 {
            replicas++
        } else if cpu < targetCPU-20 && replicas > 1 {
            replicas--
        }
        fmt.Printf("tick=%02d cpu=%5.1f%% replicas=%d\n", t, cpu, replicas)
        time.Sleep(150 * time.Millisecond)
    }
}

Step-by-Step Explanation

1. Model desired and actual state explicitly.
2. Compute a minimal diff.
3. Apply one idempotent reconciliation step.
4. Observe updated state and repeat.
5. Keep loops convergent and failure-aware.

Code Anatomy

• State model captures control-plane truth.
• Reconcile function decides next action.
• Loop executes continuously with visibility into actions.

Learning Goals

• Learn the controller pattern used in modern platforms.
• Build robust automation through repeated reconciliation.
• Understand convergence over one-shot scripting.