Skip to content

Deployments

A Deployment is the most commonly used workload resource in Kubernetes.

Deployments manage the lifecycle of applications by controlling:

  • pod creation
  • scaling
  • rolling updates
  • rollbacks

Why Deployments Exist

Running pods directly would require manual management.

Deployments automate this process by ensuring the desired number of pods are always running.

If a pod crashes, Kubernetes automatically replaces it.

Deployment Architecture

flowchart TD Deployment["Deployment"] ReplicaSet["ReplicaSet"] Pod1["Pod"] Pod2["Pod"] Pod3["Pod"] Deployment --> ReplicaSet ReplicaSet --> Pod1 ReplicaSet --> Pod2 ReplicaSet --> Pod3

When a Deployment is created:

  1. Kubernetes creates a ReplicaSet
  2. The ReplicaSet creates Pods
  3. The ReplicaSet ensures the correct number of pods remain running

Rolling Updates

One of the most powerful features of Deployments is rolling updates.

flowchart LR OldPods["Old Version Pods"] NewPods["New Version Pods"] OldPods --> NewPods

This allows applications to update gradually without downtime.

During a rolling update:

  • new pods are started
  • old pods are terminated gradually

This ensures the application remains available.

Rollbacks

If a deployment fails, Kubernetes allows quick rollback to a previous version.

This makes deployments safer and reduces downtime during application updates.

Key Takeaway

Deployments provide a declarative way to manage application updates and scaling in Kubernetes.