• short intro
  • aws architect said to use autoscaling groups
    • this means we need a new deployment strategy
  • get everything into terraform and AWS
    • ci/cd pipeline
    • logging
    • notifications
  • started again using 3rd party modules
    • X lines of vpc hcl replaced with Y
  • separated internal modules so they could be reused
    • infra and web for staging and production
  • conclusion
    • terraform is slow now
    • want to break it up into separate sections that run independently