AWS

This guide will show you how to provision (Create, Update, Delete) an AWS S3 Bucket using Kubeform.

Examples used in this guide can be found here.

At first, let’s look at the Terraform configuration for an AWS S3 Bucket below:

provider "aws" {
    access_key = "ACCESS_KEY"

    region = "REGION_NAME"

    secret_key = "SECRET_KEY"
}

resource "aws_s3_bucket" "test1" {
    bucket = "s3-bucket-test1"
}

Now, if we apply terraform apply this config will create an AWS S3 Bucket. We’ll create the exact configuration using kubeform. The steps are given below:

1. Before You Begin

At first, you need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. If you do not already have a cluster, you can create one by using kind.

Now, install Kubeform AWS provider operator in your cluster following the steps here. To get a FREE license, please visit here.

helm install kubeform-provider-aws appscode/kubeform-provider-aws \
  --namespace kubeform --create-namespace \
  --set-file kubeform-provider.license=/path/to/the/license.txt \
  --set crds.s3=true

To keep things isolated, this tutorial uses a separate namespace called demo throughout this tutorial.

$ kubectl create ns demo
namespace/demo created

2. Create AWS Provider Secret

Then create the secret which is necessary for provisioning the S3 Bucket in AWS.

apiVersion: v1
kind: Secret
metadata:
  name: aws-provider-secret
  namespace: demo
stringData:
  provider: |
    {
        "region": "<REGION_NAME>",
        "access_key": "<ACCESS_KEY>",
        "secret_key": "<SECRET_KEY>"
    }    

Here we can see that, the provider field of the stringData of the secret is same as the field of the provider part in the terraform config file. The provider secret needs to be provided in json format, under the provider key. Save it in a file (eg. provider-secret.yaml) then apply it using kubectl.

kubectl apply -f provider-secret.yaml

Note: Here, key of the provider field of the stringData (eg. "access_key", "secret_key" ) must be in snake case format (same as the tf configuration file)

3. Create S3 Bucket

Now, we’ll create the S3 Bucket CRD. The yaml is given below:

apiVersion: s3.aws.kubeform.com/v1alpha1
kind: Bucket
metadata:
  name: test1
  namespace: demo
spec:
  resource:
    bucket: s3-bucket-test1
  providerRef:
    name: aws-provider-secret
  terminationPolicy: DoNotTerminate

Here, the resource field contains the AWS S3 Bucket resource spec. Also, we can see that the provider secret is referenced using a field called providerRef.

We can see a field named terminationPolicy, this is a feature of Kubeform. This field can have two values, Delete or DoNotTerminate. When the value of this field is set to DoNotTerminate then the resource won’t get deleted even though we apply kubectl delete operation, this field needs to be set to Delete to delete the resource. It helps to avoid accidental deletion of the resource. We will see the use of this field in Delete S3 Bucket part later on this guide.

Save it in a file (eg. aws-s3-bucket.yaml) then apply it using kubectl.

kubectl apply -f aws-s3-bucket.yaml

After applying this command, the resource will be in InProgress phase until the cloud creates the resource. Once the cloud resource get created, the resource will be in Current phase which means we have successfully created the resource.

After successful creation of the resource, the resource state is available under spec.state section. This spec.state field maps the real world resource to the Kubeform resource. This field doesn’t contain any sensitive field. Sensitive fields are stored in the secret specified in the spec.secretRef section. If no secretRef is specified, Kubeform will create one.

4. Update S3 Bucket

Now, we’ll update the S3 Bucket CRD. For updating the Bucket, we will modify the existing aws-s3-bucket.yaml, we will use a different bucket (s3-bucket-test1-update) field. The modified yaml is given below:

apiVersion: s3.aws.kubeform.com/v1alpha1
kind: Bucket
metadata:
  name: test1
  namespace: demo
spec:
  resource:
    bucket: s3-bucket-test1-update
  providerRef:
    name: aws-provider-secret
  terminationPolicy: DoNotTerminate

Now, apply it using kubectl command.

kubectl apply -f aws-s3-bucket.yaml

After that, existing AWS S3 Bucket will be updated!

Note: Here, we have changed the bucket field which is Immutable, means if we change an immutable field then the resource will first get deleted and then created. But, there are some fields which are mutable, means changing those fields, resource will be only updated/changed. So, be careful!

5. Delete S3 Bucket

To delete the S3 Bucket just run:

kubectl delete -f aws-s3-bucket.yaml

After applying this command we will get below error message, as we have set terminationPolicy: DoNotTerminate:

Error from server (bucket "default/test1" can't be terminated. To delete, change spec.terminationPolicy to Delete): error when deleting "aws-s3-bucket.yaml": admission webhook "bucket.s3.aws.kubeform.com" denied the request: bucket "default/test1" can't be terminated. To delete, change spec.terminationPolicy to Delete

Let’s change the terminationPolicy to Delete by using kubectl patch command.

kubectl patch -n demo bucket test1 -p '{"spec":{"terminationPolicy":"Delete"}}' --type="merge"

Now, we can delete the Bucket.

kubectl delete -f aws-s3-bucket.yaml

After applying this command the resource will be in Terminating phase until the cloud resource get destroyed. Once the cloud resource get destroyed, the resource will get deleted successfully.