Project goal

This simple project demonstrates the conditional creation of resources using Pulumi, based on the state of pre-existing resources.

Inspired by this blog post.

Setup

First of all clone the repo:

git clone https://github.com/wtfzambo/pulumi-conditional-resource.git

Creating resources

You will need to create 2 AWS SSM parameters and 1 bucket. Open the AWS console and create the following resources:

• SSM parameter 1: /pulumiExample/dev/shouldCreateBucket with value true
• SSM parameter 2: /pulumiExample/prod/shouldCreateBucket with value false
• S3 bucket: pulumiexample-prod-<YOUR_AWS_ACCOUNT_ID>

You can quickly retrieve the AWS account ID by running the following command if you have AWS CLI installed:

aws sts get-caller-identity

Environment

This project uses named AWS profiles for authentication, so make sure you have the right setup beforehand.

Export the AWS_PROFILE environment variable:

• export AWS_PROFILE=my_profile on Mac or Linux
• $env:AWS_PROFILE=my_profile on Windows Powershell

Install pulumi

• brew install pulumi on Mac or Linux
• scoop install pulumi on Windows

Pulumi can use different backends to manage state. For simplicity, let's use a local one.

pulumi login --local

Install python packages

python -m venv .venv
source .venv/bin/activate  # (.venv\Scripts\activate on Windows)
pip install -r requirements.txt

Running the project

Pulumi has the concept of stacks which can be compared to environments. Select the dev stack: pulumi stack select dev and then run:

pulumi preview

This will show you the resources that are about be created. You will notice that in the outputs, the ARN of the bucket is not displayed, showing output<string> instead. This is because the resource doesn't exist yet.

Now run

pulumi up -y

Pulumi has created the resources and the ARN of the bucket is now available in the output.

---

Now select the prod stack: pulumi stack select prod and then run the same commands as above. In this case, you will see that even during preview, the ARN of the bucket is available. This is because on the condition we set, the prod stack shouldn't create a new bucket, but just reference the already existing one that you created in the first steps.