The main difference between kubectl apply -f and kubectl create -f is the way of resource processing. 1. create is used to create a new resource, and if it already exists, an error will be reported; 2. apply it automatically determines whether the resource exists and chooses to create or update; 3. create is suitable for first deployment, avoiding overriding configuration or script automation, apply is suitable for scenarios where iterative modification, infrastructure is code, and uncertain resource status; 4. apply is more flexible and can intelligently handle field differences, but resources such as ConfigMap and Secret need to be used with --dry-run; 5. CRD usually needs to use create, while Helm or Kustomize internally depends on apply to manage state and differences. Although both can deploy YAML files, the applicable scenarios are different.

The main difference between kubectl apply -f and kubectl create -f lies in how they handle the creation and updating of Kubernetes resources.

Creation vs. Apply Strategy

When you use kubectl create -f , it expects that the resource doesn't already exist. If it does, you'll get an error saying the resource already exists. This makes create a one-time operation — great for fresh deployments.

On the other hand, kubectl apply -f uses a declarative approach. It looks at the current state of the resource in the cluster and figures out what needs to be changed to match the configuration in your file. So if the resource doesn't exist, it creates it; if it does exist, it updates only the necessary fields.

This makes apply more flexible, especially when managing environments like staging or production where config changes happen over time.

Use Cases: When to Use Which

Here are some typical scenarios where you'd prefer one command over the other:

• Use kubectl create -f when :

  • You're setting up something for the first time.
  • You want to avoid accidentally overwriting existing configurations.
  • You're scripting automation that should fail fast if a resource already exists.

• Use kubectl apply -f when :

  • You're making iterative changes to your deployment files.
  • Managing infrastructure as code (like GitOps workflows).
  • You don't want to worry about whether a resource already exists.

One common pitfall is using create repeatedly during development, which leads to errors when trying to recreate existing resources. In those cases, people often turn to delete followed by create , but that's more steps than needed — apply handles it all in one go.

Behavior with Complex Resources

Some Kubernetes resources have complex behaviors when created versus applied. For example:

• Services and Deployments usually behave similarly under both commands.
• ConfigMaps and Secrets can cause issues if you apply them without using the --dry-run=client -o yaml | kubectl apply -f - pattern to manage immutable fields properly.
• CustomResourceDefinitions (CRDs) typically requires create because applying them may not work consistently across versions.

Also, when using tools like Helm or Kustomize, they tend to rely on apply internally because they manage state and diffs themselves.

So while both commands can deploy resources from YAML files, create is more rigid and strict, while apply is smarter and more forgiving for day-to-day operations.

Basically that's it.

The above is the detailed content of What is the difference between kubectl apply -f and kubectl create -f?. For more information, please follow other related articles on the PHP Chinese website!