Design a Deployment on Kubernetes

Design a Deployment on Kubernetes

Step 1 of 13

The brief: a checkout API that stays up, scales with traffic, and updates safely. Start at the front door.

The Idea

The interview answer is a walk along the request path, then through the failures. An Ingress or Gateway routes to a Service, the Service to ready Pods of a Deployment spread across zones. Config comes from ConfigMaps and Secrets, state lives in a StatefulSet or a managed database, and every container declares requests, limits and probes.

Real-World Example

"Deploy a checkout API": at least three replicas spread by zone, an HPA on CPU, readiness on /ready, a PodDisruptionBudget of minAvailable: 2 so node drains never evict too many at once, and rolling updates with maxUnavailable: 0.

The Tradeoff

Every safety net costs something: spare capacity for surges, extra zones for spread. A PodDisruptionBudget covers only voluntary disruptions such as drains; a node crash still counts against it. Say which failure each piece handles.

Hands-On

# illustrative — two guards a checkout API adds
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata: { name: checkout }
spec:
  minAvailable: 2
  selector: { matchLabels: { app: checkout } }
---
# in the Deployment's Pod template
topologySpreadConstraints:
  - maxSkew: 1
    topologyKey: topology.kubernetes.io/zone
    whenUnsatisfiable: DoNotSchedule
    labelSelector: { matchLabels: { app: checkout } }

Your turn

Put the steps in the right order.

1. 1The Service picks a ready checkout Pod↑↓
2. 2The Ingress controller routes /checkout to the checkout Service↑↓
3. 3The user's request reaches the cluster's load balancer↑↓
4. 4The Pod reads its database password from a mounted Secret and queries the database↑↓

Check order

Mini quiz

1 / 3

A PodDisruptionBudget with minAvailable: 2 protects against:

• ANode hardware failure
• BA bad image in a rollout
• CToo many Pods evicted at once by voluntary disruptions such as node drains
• DDNS outages

Sources

• Disruptions and PodDisruptionBudgets — Kubernetes Documentation
• Pod topology spread constraints — Kubernetes Documentation
• Horizontal Pod Autoscaling — Kubernetes Documentation
• Deployments — Kubernetes Documentation