Looking for some guidance on configuring application stages

[andrewgadziksonos]

Hello!

I'm currently looking for some guidance around how to properly configure application stages (dev, test, perf, etc.). I currently have access to 3 AWS accounts,

1. AWS-Prod
2. AWS-Dev
3. AWS-Sandbox

I have strict orders from our IT department to only put production infrastructure in AWS-Prod, and "non prod" infrastructure in AWS-Dev (Sandbox is only used for testing CDK configurations, etc).

Our team wants to make use of CDK Pipelines to automate most of our CI/CD needs. Our plan is trigger a pipeline after a code push into the main branch. The following steps are our pipeline stages,

• test code
• build code
• deploy to DEV
• await manual approval
• deploy to TEST
• run e2e tests
• deploy to STAGE
• run smoke tests
• await manual approval
• deploy to PRODUCTION

I assume that I'll have some sort of ProdAppStack and NonProdAppStack defined along with my PipelineStack

With all of that said, my initial thought is to contain our pipeline in our AWS-Prod account. Does this sound accurate to everyone?

If so, I assume I'm able to build out the "non-prod" infra from the AWS-Prod account? Is there any sort of special configuration I need to do beforehand for our AWS-Prod account to gain access to AWS-Dev?

Thanks in advance for the help!

-Andrew

[tim-finnigan]

Hi @andrewgadziksonos thanks for your patience. It has been a while since this discussion was created — did you have any updates on your end?

I think this documentation on continuous integration and delivery (CI/CD) using CDK Pipelines will help as a reference for what you’re trying to do. There is also an example here in the CDK Workshop on configuring pipelines with various stages using the CDK. Please let us know if that helps or if you had any follow up questions.

[github-actions[bot]]

Hello! Reopening this discussion to make it searchable.