Deployment managers

Introduction #

In Kubernetes, workload controllers orchestrate and automate the life cycle of containerized applications. They handle scaling, redundancy and resource management for applications, enabling a scalable, fault-tolerant architecture.

Deployment #

Controller for stateless applications.

Overview #

A Deployment deploys and manages stateless applications. It maintains the desired number of running Pods, handles updates and rollbacks, and offers horizontal scaling options.

Use cases #

Used for web applications, APIs or any stateless service that can easily be replicated without depending on persistent storage.

Example #

apiVersion: apps/v1
kind: Deployment
metadata:
  name: example-deployment
spec:
  replicas: 3
  selector:
    matchLabels:
      app: example
  template:
    metadata:
      labels:
        app: example
    spec:
      containers:
      - name: example
        image: nginx

ReplicaSet #

Maintains the number of Pod replicas.

Overview #

A ReplicaSet maintains a constant number of identical Pods in the cluster. Deployments often rely on ReplicaSets to create and update Pods.

Use cases #

Used internally by Deployments to keep a stable number of Pods running.

StatefulSet #

For applications that need persistent storage.

Overview #

A StatefulSet is used for applications that need data persistence. Each Pod has a unique identity, preserved when the Pod is recreated, and is associated with a dedicated storage volume.

Use cases #

Databases, distributed file systems, or any application that needs a unique identifier or persistent state.

Example #

apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: example-statefulset
spec:
  serviceName: "example"
  replicas: 3
  selector:
    matchLabels:
      app: example
  template:
    metadata:
      labels:
        app: example
    spec:
      containers:
      - name: example
        image: mysql

DaemonSet #

Ensures a Pod runs on every node.

Overview #

A DaemonSet ensures that a Pod runs on every node in the cluster. It is used for monitoring, log analysis or network maintenance tasks.

Use cases #

Monitoring agents, log collection, and any service that needs to be present on every node.

Example #

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: example-daemonset
spec:
  selector:
    matchLabels:
      name: example
  template:
    metadata:
      labels:
        name: example
    spec:
      containers:
      - name: example
        image: fluentd

Job #

Running one-off tasks.

Overview #

A Job runs Pods to complete a one-off task and stops when the task is finished. It is especially suited to data processing or maintenance scripts.

Use cases #

One-off tasks such as data analysis or batch processing.

Example #

apiVersion: batch/v1
kind: Job
metadata:
  name: example-job
spec:
  template:
    spec:
      containers:
      - name: example
        image: busybox
        command: ["echo", "Hello, Kubernetes!"]
      restartPolicy: Never

CronJob #

Scheduling recurring tasks.

Overview #

A CronJob schedules recurring tasks using crontab syntax. It is useful for backups, report generation or any periodic task.

Use cases #

Recurring tasks such as backups or data cleanup.

Example #

apiVersion: batch/v1beta1
kind: CronJob
metadata:
  name: example-cronjob
spec:
  schedule: "*/5 * * * *"
  jobTemplate:
    spec:
      template:
        spec:
          containers:
          - name: example
            image: busybox
            command: ["echo", "Hello from CronJob"]
          restartPolicy: OnFailure

Choosing a workload controller #

• For stateless applications: use a Deployment.
• For stateful applications that need persistent identifiers: prefer a StatefulSet.
• For tasks that must run on every node: choose a DaemonSet.
• For one-off tasks: a Job is recommended.
• For recurring tasks: a CronJob is ideal.