Skip to content

Latest commit

 

History

History
58 lines (45 loc) · 2.12 KB

readme.md

File metadata and controls

58 lines (45 loc) · 2.12 KB
Raw

Provisioning BRAVO on AWS

This terraform project is designed to provision infrastructure that will run the BRAVO API and UI. Major components created:

  • EC2 instance running the API
  • EC2 instance for running mongodb
  • EBS mounts for API and mongo instances.
  • Bucket for storing the Vue UI
  • Cloudfront distribution for serving out the UI.

Running this provisioner will incur charges to the associated AWS account.

Dependencies

  • AWS access key and access secret for deployment
  • Terraform cloud project
  • Existing Domain registered with Route53 e.g. example.com
  • A cert for that domain that with a SAN that covers bravo subdomain in SANs.
    • Expected cert name is the top level name e.g. example.com
    • e.g Cert with name example.com having *.example.com in SAN list.
  • Backing vignette data in an S3 bucket.
    • Available from: ftp://share.sph.umich.edu/bravo/bravo_vignette_data.tar.bz2
    • Backing data needs to be present in bucket.

Required Terraform Variables

Terraform variables are stored in the terraform cloud project. Full list of names and descriptions available there or in variables.tf. Some of the important ones are:

  • Name of key pair to use to access EC2 instances.
  • Names of hosted zone domain, ui domain, api domain.
  • Name of bucket that contains backing data.

Development debugging simple httpd server

The init-script.sh includes a simple httpd systemd service to indicate test port 8080. Enable running this init script as the instance user data with

terraform apply -var='install_httpd=true'

Do NOT use this when you will be deploying the BRAVO API. Deploy will fail because the port for the API to bind to is already in use.

Provisioning: Stating or Production

This code base provisions for both the staging and production infrastructure. See managing workspaces for docs. In short the relevant cli commands are:

  • terraform workspace list
  • terraform workspace select

Make sure you are on the correct workspace

With the correct tf cloud workspace selected, provision using as usual with:

terraform apply