K8s PodDisruptionBudget PDB Guide

💡 Quick Answer: PodDisruptionBudget prevents too many pods from being evicted simultaneously during voluntary disruptions (node drain, cluster upgrades). Set minAvailable: 2 (at least 2 pods must stay running) or maxUnavailable: 1 (at most 1 pod down at a time). Without a PDB, kubectl drain can evict all pods at once.

The Problem

Voluntary disruptions can kill your application:

• kubectl drain node evicts all pods on a node
• Cluster autoscaler scales down nodes with pods
• Kubernetes upgrades drain nodes one by one
• Without PDB, all replicas of a service could be evicted simultaneously

The Solution

Basic PDB

# At least 2 pods must always be running
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: web-pdb
spec:
  minAvailable: 2
  selector:
    matchLabels:
      app: web-frontend

---
# At most 1 pod can be unavailable
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: api-pdb
spec:
  maxUnavailable: 1
  selector:
    matchLabels:
      app: api-service

---
# Percentage-based
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: worker-pdb
spec:
  maxUnavailable: "25%"
  selector:
    matchLabels:
      app: worker

How PDB Works

# Without PDB:
kubectl drain node-1
# All 3 replicas on node-1 evicted simultaneously → downtime!

# With PDB (maxUnavailable: 1):
kubectl drain node-1
# Pod 1 evicted → rescheduled on node-2 → becomes Ready
# Pod 2 evicted → rescheduled on node-3 → becomes Ready
# Pod 3 evicted → rescheduled → done
# At most 1 pod unavailable at any time → no downtime!

Check PDB Status

kubectl get pdb
# NAME      MIN AVAILABLE   MAX UNAVAILABLE   ALLOWED DISRUPTIONS   AGE
# web-pdb   2               N/A               1                     5m
# api-pdb   N/A             1                 1                     5m

kubectl describe pdb web-pdb
# Status:
#   Current Healthy:   3
#   Desired Healthy:   2
#   Disruptions Allowed: 1

PDB + Node Drain

# Drain respects PDBs
kubectl drain node-1 --ignore-daemonsets --delete-emptydir-data

# If PDB blocks drain:
# error when evicting pods: Cannot evict pod as it would violate the pod's disruption budget.

# Force drain (ignores PDB — use with caution!)
kubectl drain node-1 --ignore-daemonsets --force

# Timeout drain
kubectl drain node-1 --ignore-daemonsets --timeout=300s

minAvailable vs maxUnavailable

Common Patterns

# Stateless web app — allow 1 down at a time
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: web-pdb
spec:
  maxUnavailable: 1
  selector:
    matchLabels:
      app: web

---
# Database — never disrupt the primary
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: db-pdb
spec:
  minAvailable: 1        # Primary must always be up
  selector:
    matchLabels:
      app: postgres
      role: primary

---
# Batch workers — allow more disruption
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
  name: worker-pdb
spec:
  maxUnavailable: "50%"
  selector:
    matchLabels:
      app: batch-worker

Common Issues

Drain blocked indefinitely by PDB

Only 1 replica and minAvailable: 1 — can never evict. Use maxUnavailable: 1 instead, or increase replicas.

PDB not protecting pods

Selector doesn’t match pod labels. Check: kubectl get pdb -o yaml and compare with kubectl get pods --show-labels.

PDB blocks cluster autoscaler scale-down

By design — autoscaler respects PDBs. Ensure enough replicas spread across nodes so at least one node can always be drained.

Best Practices

• Every production Deployment should have a PDB — protect against drain
• Use maxUnavailable: 1 as the default — simple and safe
• Never set minAvailable equal to replicas — blocks all disruptions
• Spread replicas across nodes with pod anti-affinity — PDB + spread = safe drains
• Test with kubectl drain --dry-run=server — preview what would be evicted

Key Takeaways

• PDBs protect pods during voluntary disruptions (drain, autoscaler, upgrades)
• maxUnavailable: 1 is the most common and safest default
• Without PDB, kubectl drain can evict all replicas simultaneously
• PDBs don’t protect against involuntary disruptions (node crash, OOM)
• Combine with pod anti-affinity for full high-availability protection