World map

 

BM: Baca dalam Bahasa Melayu

World K5: Lunch Rush · Boss challenge

Lunch rush

You will learn: to survive a traffic spike with an HPA, then climb the debug ladder to find what slows the shop.

Checked with Kubernetes 1.37 on 31 Oct 2026

The problem

11:55. The food court opens in five minutes. Your shop has 3 trays and metrics-server. Boss mission: survive lunch, then find what slows it down. What do you set up first?

Pick one answer.

Try it

Room 1: make the HPA. Then press Lunch rush. Survive with at least 6 ready trays.

You are here: cluster kind-dev namespace default
Stalls: 3. Trays: 3 (3 ready). Waiting for a stall: 0.
  • shop3/3 ready
dev-control-planeOpenManager

No trays here.

dev-workerOpen
  • shop·ch69s
dev-worker2Open
  • shop·59vn4
  • shop·rg4hp
Time: +0s (sped up)

Predict

Next day, the HPA asks for 10 trays. Only 6 start. 4 stay Pending. What could hold them back?

Pick one answer.

Try it

Room 2: before you climb, check your ladder. Bottom rung at the top.

Bottom rung at the top. Put these in order. Use the arrows, or drag.

  • debug
  • get
  • describe
  • logs
  • events
  • exec

Try it

Room 3: the rush is on, and 4 trays wait. Climb the ladder. Find who holds the stalls, then free them. Goal: 10 ready shop trays.

You are here: cluster kind-dev namespace default
Stalls: 3. Trays: 8 (8 ready). Waiting for a stall: 4.
  • shop6/10 ready
dev-control-planeOpenManager

No trays here.

dev-workerOpen
  • loadtest·xhzwz
  • shop·bwzjl
  • shop·l4j5r
  • shop·q4rvd
dev-worker2Open
  • loadtest·hhg79
  • shop·dttjb
  • shop·g48pw
  • shop·rcq4b
Waiting for a stall
  • shop·4f482Pending
  • shop·9ckwxPending
  • shop·tmpq9Pending
  • shop·zrt48Pending
Time: +0s (sped up)

Check

kubectl top showed loadtest using about 5.7Gi each. But its YAML requests 7Gi each. Which number blocked the shop trays?

Pick one answer.

Check

Plan for tomorrow’s lunch. Good idea, or needs work?

Swipe right for good, left for needs work. Or use the buttons.

1 / 4

Run load tests in their own zone, with a ResourceQuota

Run load tests in their own zone, with a ResourceQuota

You did it!

You can now survive a lunch rush and find what slows it down.

Go deeper (optional)

Real rep (optional), on a throwaway kind cluster with metrics-server. Start a Service called shop, then make a load tray. Delete it when you are done.

kubectl run load --image=busybox:1.37 --restart=Never -- /bin/sh -c "while sleep 0.01; do wget -q -O- http://shop; done"
kubectl get hpa shop --watch
kubectl delete pod load