Overview

Cloud infrastructure changes rapidly, and often the manual nature of infrastructure-as-code workflows can become a drag on velocity, causing organizations to reach for more self-service approaches. This can involve stitching together Backstage, GitLab, ArgoCD, or any other set of tools. Plural provides a single cloud orchestrator to provide all those key functionalities in one control plane.

We'll show how this can work beyond a Kubernetes provisioning use case, to provisioning an S3 bucket, a common resource also needed by Kubernetes workloads. It'll operate by:

• Creating a PR Automation (PRA) to make the provisioning of buckets repeatable
• Using that PRA to create an InfrastructureStack to provision the S3 bucket using a Plural Terraform stack.

Define the PR Automation

This PR Automation is going to leverage resources that have already been created in the plural up repo for you, in particular:

• terraform/modules/blob/s3 - a basic S3 bucket module we've predefined that can be used for provisioning the bucket
• templates/blobstore/* - a couple of templates that will be used to set up the blobstore provisioning process via a PRA

Given those manifests, the PR Automation YAML could be something like this, which you'd place in bootstrap/pr-automation/blobstore.yaml:

apiVersion: deployments.plural.sh/v1alpha1
kind: PrAutomation
metadata:
  name: blob-creator
spec:
  name: blob-creator
  documentation: |
    Sets up a PR to provision a blobstore with a given type (eg s3) and region
  creates:
    templates:
    - source: templates/blob/stack.yaml
      destination: "services/blobstores/{{ context.type }}/{{ context.name }}.yaml"
      external: false
    - source: templates/blob/service.yaml
      destination: "bootstrap/blobstores.yaml"
      external: false
  scmConnectionRef:
    name: github
  title: "Adding a {{ context.type }} bucket {{ context.name }}"
  message: "Setup a stack to manage the {{ context.name }} {{ context.type }} bucket"
  identifier: your-org/your-plural-up-repo # <---- replace with the slug for your plural up repo
  configuration:
  - name: name
    type: STRING
    documentation: the name of this blob store (if using s3, this would become an s3 bucket name)
    validation:
      regex: "[a-z][a-z-0-9]+"
  - name: type
    type: ENUM
    documentation: the type of blob storage to provision
    values:
    - s3
  - name: region
    type: STRING
    documentation: the region your blobstore will live in