Workloads

Understanding Kubernetes workload resources: ReplicaSet, Deployment, DaemonSet, StatefulSet, Job, CronJob

Overview

Workload resources manage the lifecycle of Pods. Kubernetes provides several workload controllers, each designed for specific use cases. This section covers the different workload types and when to use each one.

Workload Comparison

Workload	Use Case	Scaling	Storage	Updates
ReplicaSet	Maintain N identical pods	Manual	No	No
Deployment	Stateless apps	Yes (auto)	No	Yes (rolling)
StatefulSet	Stateful apps	Yes (auto)	Yes (unique)	Yes (rolling)
DaemonSet	One pod per node	N/A	No	Yes (rolling)
Job	Run to completion	No	Optional	No
CronJob	Scheduled jobs	No	Optional	No

Study Path

ReplicaSet - Maintain N identical pods
Deployment - Stateless applications
DaemonSet - One pod per node
StatefulSet - Stateful applications
Job - Run to completion
CronJob - Scheduled jobs

Quick Reference

Choosing a Workload Type

┌─────────────────────────────────────────────────────────────────┐
│                    Need to Deploy Pods?                        │
└─────────────────────────────────────────────────────────────────┘
                            ↓
        ┌───────────────────────┬───────────────────────┐
        │ Run once to completion? │                      │
        │ YES → Job/CronJob       │ NO                   │
        └───────────────────────┴───────────────────────┘
                            ↓
        ┌───────────────────────┬───────────────────────┐
        │ One pod per node?      │                       │
        │ YES → DaemonSet        │ NO                    │
        └───────────────────────┴───────────────────────┘
                            ↓
        ┌───────────────────────┬───────────────────────┐
        │ Needs stable identity? │                       │
        │ + Persistent storage?  │                       │
        │ YES → StatefulSet      │ NO → Deployment       │
        └───────────────────────┴───────────────────────┘

Common Commands

bash

# Deployments
kubectl create deployment nginx --image=nginx
kubectl scale deployment nginx --replicas=3
kubectl set image deployment/nginx nginx=nginx:1.26
kubectl rollout undo deployment/nginx

# StatefulSets
kubectl get statefulsets
kubectl scale statefulset mongodb --replicas=5
kubectl rollout status statefulset/mongodb

# DaemonSets
kubectl get daemonsets
kubectl rollout restart daemonset/fluentd

# Jobs
kubectl create job my-job --image=busybox -- echo "Hello"
kubectl create cronjob my-cron --image=busybox --schedule="*/5 * * * *"
kubectl get jobs
kubectl delete job my-job

Workload Specifications

ReplicaSet

yaml

apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: nginx-replicaset
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.25

Deployment

yaml

apiVersion: apps/v1
kind: Deployment
metadata:
  name: nginx-deployment
spec:
  replicas: 3
  selector:
    matchLabels:
      app: nginx
  template:
    metadata:
      labels:
        app: nginx
    spec:
      containers:
      - name: nginx
        image: nginx:1.25
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 0

DaemonSet

yaml

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: fluentd
spec:
  selector:
    matchLabels:
      app: fluentd
  template:
    metadata:
      labels:
        app: fluentd
    spec:
      containers:
      - name: fluentd
        image: fluentd:v1.16

StatefulSet

yaml

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: mongodb
spec:
  serviceName: mongodb
  replicas: 3
  selector:
    matchLabels:
      app: mongodb
  template:
    metadata:
      labels:
        app: mongodb
    spec:
      containers:
      - name: mongodb
        image: mongo:6.0
        ports:
        - containerPort: 27017
  volumeClaimTemplates:
  - metadata:
      name: mongodb-data
    spec:
      accessModes: ["ReadWriteOnce"]
      resources:
        requests:
          storage: 10Gi

Job

yaml

apiVersion: batch/v1
kind: Job
metadata:
  name: batch-job
spec:
  completions: 1
  backoffLimit: 4
  template:
    spec:
      restartPolicy: OnFailure
      containers:
      - name: worker
        image: busybox
        command: ["sh", "-c", "echo 'Hello World'"]

CronJob

yaml

apiVersion: batch/v1
kind: CronJob
metadata:
  name: scheduled-job
spec:
  schedule: "0 */6 * * *"  # Every 6 hours
  successfulJobsHistoryLimit: 3
  failedJobsHistoryLimit: 3
  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: OnFailure
          containers:
          - name: cron-worker
            image: busybox
            command: ["sh", "-c", "echo 'Running scheduled job'"]

Common Patterns

Stateless Application (Deployment)

yaml

# Web servers, APIs, microservices
apiVersion: apps/v1
kind: Deployment
metadata:
  name: web-api
spec:
  replicas: 3
  selector:
    matchLabels:
      app: web-api
  template:
    metadata:
      labels:
        app: web-api
    spec:
      containers:
      - name: api
        image: my-api:v2.1
        resources:
          requests:
            memory: "256Mi"
            cpu: "250m"
          limits:
            memory: "512Mi"
            cpu: "500m"

Stateful Application (StatefulSet)

yaml

# Databases, message queues, distributed systems
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: postgres-db
spec:
  serviceName: postgres-db
  replicas: 3
  selector:
    matchLabels:
      app: postgres
  template:
    metadata:
      labels:
        app: postgres
    spec:
      containers:
      - name: postgres
        image: postgres:14
        env:
        - name: POSTGRES_PASSWORD
          value: "secret123"
        ports:
        - containerPort: 5432
        volumeMounts:
        - name: postgres-data
          mountPath: /var/lib/postgresql/data
  volumeClaimTemplates:
  - metadata:
      name: postgres-data
    spec:
      accessModes: ["ReadWriteOnce"]
      resources:
        requests:
          storage: 10Gi

Batch Processing (Job)

yaml

# ML training, data processing, backups
apiVersion: batch/v1
kind: Job
metadata:
  name: ml-training-job
spec:
  completions: 1
  parallelism: 1
  backoffLimit: 4
  activeDeadlineSeconds: 3600  # 1 hour max
  template:
    spec:
      restartPolicy: Never
      containers:
      - name: trainer
        image: ml-trainer:latest
        command: ["python", "train.py"]
        resources:
          limits:
            nvidia.com/gpu: 1

Cluster-Level Agent (DaemonSet)

yaml

# Monitoring agents, log collectors, network plugins
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: monitoring-agent
spec:
  selector:
    matchLabels:
      app: monitoring-agent
  template:
    metadata:
      labels:
        app: monitoring-agent
    spec:
      containers:
      - name: agent
        image: monitoring-agent:v1.0
        resources:
          requests:
            memory: "64Mi"
            cpu: "50m"
          limits:
            memory: "128Mi"
            cpu: "100m"

Scheduled Task (CronJob)

yaml

# Periodic backups, reports, maintenance tasks
apiVersion: batch/v1
kind: CronJob
metadata:
  name: daily-backup
spec:
  schedule: "0 2 * * *"  # 2 AM daily
  concurrencyPolicy: Forbid  # Don't overlap
  successfulJobsHistoryLimit: 7
  failedJobsHistoryLimit: 3
  jobTemplate:
    spec:
      template:
        spec:
          restartPolicy: OnFailure
          containers:
          - name: backup
            image: backup-tool:latest
            command: ["./backup.sh"]

Update Strategies

RollingUpdate (Deployment, DaemonSet)

yaml

spec:
  strategy:
    type: RollingUpdate
    rollingUpdate:
      maxUnavailable: 25%  # Max 25% can be down during update
      maxSurge: 25%        # Max 25% extra pods during update

RollingUpdate (StatefulSet)

yaml

spec:
  updateStrategy:
    type: RollingUpdate
    rollingUpdate:
      partition: 0  # Update pods with ordinal >= partition

OnDelete (Manual Control)

yaml

spec:
  updateStrategy:
    type: OnDelete  # Only update when pod is deleted

Scaling

Manual Scaling

bash

kubectl scale deployment nginx --replicas=10
kubectl scale statefulset mongodb --replicas=5

Horizontal Pod Autoscaler (HPA)

yaml

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: nginx-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: nginx
  minReplicas: 3
  maxReplicas: 20
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 70

Best Practices

1. Use Deployments for Stateful Apps Only When Necessary

Deployments: Simpler, faster, better for most cases
StatefulSets: Required only when you need:
- Stable network identities
- Stable persistent storage
- Ordered deployment and scaling

2. Always Set Resource Limits

yaml

resources:
  requests:
    memory: "128Mi"
    cpu: "100m"
  limits:
    memory: "256Mi"
    cpu: "200m"

3. Use Probes for Self-Healing

yaml

livenessProbe:
  httpGet:
    path: /health
    port: 8080
  initialDelaySeconds: 30
readinessProbe:
  httpGet:
    path: /ready
    port: 8080
  initialDelaySeconds: 10

4. Pin Image Versions

yaml

# Bad
image: myapp:latest

# Good
image: myapp:v2.1.0

5. Set Appropriate Restart Policies

Workload	Recommended Policy
Deployment	Always (default)
DaemonSet	Always (default)
StatefulSet	Always (default)
Job	OnFailure or Never
CronJob	OnFailure or Never

Next Steps

Learn ReplicaSet: ReplicaSet
Understand Deployment: Deployment
Study StatefulSet: StatefulSet
Explore Jobs: Job
Practice: Lab 02: First Deployment

Continue Learning:

Practice: Lab 02: First Deployment

Return to: Key Concepts | Overview

Workloads ​

Overview ​

Workload Comparison ​

Study Path ​

Quick Reference ​

Choosing a Workload Type ​

Common Commands ​

Workload Specifications ​

ReplicaSet ​

Deployment ​

DaemonSet ​

StatefulSet ​

Job ​

CronJob ​

Common Patterns ​

Stateless Application (Deployment) ​

Stateful Application (StatefulSet) ​

Batch Processing (Job) ​

Cluster-Level Agent (DaemonSet) ​

Scheduled Task (CronJob) ​

Update Strategies ​

RollingUpdate (Deployment, DaemonSet) ​

RollingUpdate (StatefulSet) ​

OnDelete (Manual Control) ​

Scaling ​

Manual Scaling ​

Horizontal Pod Autoscaler (HPA) ​

Best Practices ​

1. Use Deployments for Stateful Apps Only When Necessary ​

2. Always Set Resource Limits ​

3. Use Probes for Self-Healing ​

4. Pin Image Versions ​

5. Set Appropriate Restart Policies ​

Next Steps ​

Workloads

Overview

Workload Comparison

Study Path

Quick Reference

Choosing a Workload Type

Common Commands

Workload Specifications

ReplicaSet

Deployment

DaemonSet

StatefulSet

Job

CronJob

Common Patterns

Stateless Application (Deployment)

Stateful Application (StatefulSet)

Batch Processing (Job)

Cluster-Level Agent (DaemonSet)

Scheduled Task (CronJob)

Update Strategies

RollingUpdate (Deployment, DaemonSet)

RollingUpdate (StatefulSet)

OnDelete (Manual Control)

Scaling

Manual Scaling

Horizontal Pod Autoscaler (HPA)

Best Practices

1. Use Deployments for Stateful Apps Only When Necessary

2. Always Set Resource Limits

3. Use Probes for Self-Healing

4. Pin Image Versions

5. Set Appropriate Restart Policies

Next Steps