Skip to content

flibbertigibbet/cac-tripplanner

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Clean Air Council Circuit Trip Planner and Travelshed

Development Dependencies

Development Installation

  1. Make sure you have the development dependencies installed
  2. Place GTFS .zip files, OSM files, and elevation .tif files (optional) in the otp_data folder
  3. Copy deployment/ansible/group_vars/development_template to deployment/ansible/group_vars/development
  4. Change into the src/ folder and run npm install to install the node modules on the host machine
  5. Run vagrant up
  6. See the app at http://localhost:8024! See OpenTripPlanner at http://localhost:9090.
  7. Running ./scripts/serve-js-dev.sh on the host will rebuild the front-end app on file change (the browser must be reloaded manually to pick up the change). Alternatively, cd /opt/app/src && npm run gulp-development can be run manually in the VM to pick up changes to the static files.

Note that if there is an existing build Graph.obj in otp_data, vagrant provisioning in development mode will not attempt to rebuild the graph, but will use the one already present.

Building AMIs

  1. Make a production group_vars file (similarly to how is described above with development). Make sure production is set to true, and also specify an app_username, which should be set to: ubuntu
  2. If building the otp machine, make sure the latest GTFS are in otp_data, then build a graph when them in the development environment provisioning. This will result in a new Graph.obj file being written to otp_data.
  3. In the project directory within the app VM, run: deployment/cac-stack.py create-ami --aws-access-key-id YOUR_ACCESS_KEY --aws-secret-access-key YOUR_SECRET_KEY --aws-role-arn YOUR_ASSUMED_ROLE_ARN
  4. The previous command builds all AMIs. To only build a single AMI, run the same command, but also specify the --machine-type parameter, which may be set to one of: bastion, otp, or app.

Launching AWS Stacks

  1. Copy deployment/default_template.yaml to deployment/default.yaml and edit variables
  2. In the project directory, for a set of Blue stacks in the Production environment, run: deployment/cac-stack.py launch-stacks --stack-type prod --stack-color blue --aws-access-key-id YOUR_ACCESS_KEY --aws-secret-access-key YOUR_SECRET_KEY --aws-role-arn YOUR_ASSUMED_ROLE_ARN
  3. The previous command will do the following:
  • Ensure the VPC stack is up in Production -- it will be launched if it isn't already running
  • Ensure the DataPlane stack is up in Production -- it will be launched if it isn't already running
  • Ensure the OtpServer Blue stack is up in Production -- it will be launched if it isn't already running
  • Ensure the WebServer Blue stack is up in Production -- it will be launched if it isn't already running
  1. Note that database migrations are not automatically run. When the DataPlane is first brought up, it is necessary to manually create the app user/db and run migrations.
  2. Launching a set of Production stacks with the other color (Green), will use the same VPC and DataPlane stacks, but will create different OtpServer and WebServer stacks (if they don't already exist).

Production Blue/Green deployment

  1. Note which color is currently running in production. Use the opposite color in the following steps.
  2. Set otp_host in production group_vars to the CloudFront distribution with the desired color.
  3. Run create_ami command to build new AMIs.
  4. Update default.yaml with new AMI ids.
  5. Run launch_stacks command to launch stacks with the desired color.
  6. Test new stacks thoroughly.
  7. Switch the public DNS record of the site to point to the new WebServer ELB DNS.
  8. The stacks of the previous color may be deleted when ready.

About

Multi-modal trip planner for the Greater Philadelphia area, featuring local destinations

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • JavaScript 43.9%
  • Python 32.4%
  • HTML 12.9%
  • SCSS 10.6%
  • Shell 0.2%