Skip to content

fibonascii/cloud-automation

Repository files navigation

cloud-automation

prerequisites to automated bulid/deploy

  • Jenkins master running in the same region as we're deploying to
  • Make sure Jenkins master is able to access this repository (saved credentials)
  • Create a branch of this repo for each new environment and name it with org-region-uid. All config changes specific to this environment should be only pushed to the new branch
  • Ami images to be deployed are accessible by this account in the same region.
    • If needed, copy ami's into the same region and add permission for the receiving account # to access it
    • Replace original ami id's with their copies in desired region
  • Create a pd-cicd-devolopment key pair
  • Create IAM role cicd-admin-role with AdministratorAccess to EC2
  • Create an S3 bucket in a new account and sync pd-dev-cicd-artifacts into the new bucket. Replace pd-dev-cicd-artifacts in sceptre yaml files with the new bucket name (CoreBootStrapRepositoryS3BucketName)
  • Update official kong ami id to match desired region
  • If manifest is needed, make sure the ami id is correct
  • Replace us-east-1 in all yaml files with the desired region (same for availability zones)
  • Update WorkerNodeImageId as the one in master is located in east-1. This image is used to spin up EKS worker nodes
  • Update cloudformation/lod-rest/templates/base.py to assure uniqueness of ClientCode, ClientEnvironmentKey and matching EnvironmentRegion. ClientEnvironmentKey is alphanumerical, 6 characters max
  • Make sure to delete every incomplete environment with: sceptre delete-env ${ENV_NAME}. Otherwise there will be naming conflicts.