💡 Quick Answer: deployments

The Problem

This is a fundamental Kubernetes topic that engineers search for frequently. A comprehensive reference with production-ready examples saves hours of trial and error.

The Solution

The Zero-Downtime Checklist

apiVersion: apps/v1
kind: Deployment
metadata:
  name: web-app
spec:
  replicas: 3
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0       # ← Never remove a pod until new one is Ready
  template:
    spec:
      terminationGracePeriodSeconds: 60
      containers:
        - name: app
          image: my-app:v2
          ports:
            - containerPort: 8080

          # 1. Readiness probe — don't send traffic until ready
          readinessProbe:
            httpGet:
              path: /ready
              port: 8080
            initialDelaySeconds: 5
            periodSeconds: 5

          # 2. preStop hook — wait for LB to remove endpoint
          lifecycle:
            preStop:
              exec:
                command: ["sh", "-c", "sleep 5"]

          # 3. Resource requests — scheduler places pods correctly
          resources:
            requests:
              cpu: 100m
              memory: 128Mi

      # 4. Pod anti-affinity — spread across nodes
      affinity:
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 100
              podAffinityTerm:
                labelSelector:
                  matchLabels:
                    app: web-app
                topologyKey: kubernetes.io/hostname

  minReadySeconds: 5          # Wait 5s after Ready before continuing rollout
---
# 5. PDB — protect during node drains
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: web-pdb
spec:
  minAvailable: 2
  selector:
    matchLabels:
      app: web-app

The Five Requirements

#	Requirement	What It Does
1	Readiness probe	Prevents traffic to unready pods
2	`maxUnavailable: 0`	Never removes old pod before new is Ready
3	preStop sleep 5	Allows LB endpoint removal before shutdown
4	Handle SIGTERM	App drains in-flight requests gracefully
5	PDB	Prevents draining too many pods at once

graph LR
    A[Deploy v2] --> B[Create v2 pod]
    B --> C{Readiness probe passes?}
    C -->|Yes, after minReadySeconds| D[Remove 1 v1 pod]
    D --> E[preStop: sleep 5s]
    E --> F[SIGTERM: drain connections]
    F --> G[Pod terminated]
    C -->|Not yet| C

Frequently Asked Questions

Why do I still get 502s during deployment?

Usually missing one of: readiness probe, preStop hook, or maxUnavailable: 0. All three are required for true zero-downtime. Also ensure your app handles SIGTERM gracefully.

Best Practices

Start with the simplest configuration that meets your needs
Test changes in staging before production
Use kubectl describe and events for troubleshooting
Document your decisions for the team

Key Takeaways

This is essential Kubernetes knowledge for production operations
Follow the principle of least privilege and minimal configuration
Monitor and iterate based on real-world behavior
Automation reduces human error and improves consistency

Zero-Downtime Deployment in Kubernetes