Heroku

Apps

• LogDNA: error logging here
• Heroku Postgres: database with nightly backups
• Heroku Redis: memory storage, used for sitewide data (top nav about pages, programs, and meta data) that is accessed often with context_processors
• Mailgun: email service for password reset
• Heroku Scheduler: runs python manage.py publish_scheduled_pages every 10 minutes
• Elasticsearch: search indexing. needs a lot of love. reindex with ./manage.py update_index. Upgrade instructions here
• New Relic: app performance monitoring

Buildpacks

• Node.js: for running webpack
• Cairo: binary file, dependency for WeasyPrint
• gdkpixbug: another WeasyPrint dependency
• Python

Config Vars

variables that set manually and not by a Heroku app

• AWS_ACCESS_KEY_ID: for boto3/django-storages connection to AWS s3
• AWS_SECRET_ACCESS_KEY: or boto3/django-storages connection to AWS s3
• S3_BUCKET_NAME: name of s3 bucket to upload media content (images)
• SECRET_KEY: see docs
• STATIC_S3_BUCKET_NAME: name of s3 for static assets (logos, fonts, js/css)
• STATIC_URL: base url to use in front of /static. a cloudfront url. used in django templates and webpack to work out full url of static assets.
• CREATESEND_API_KEY: for use in syncing campaign monitor lists
• CREATESEND_CLIENTID: same as above
• CREATESEND_LISTID: same as above
• DEBUG: not used. always false in production.
• DJANGO_SETTINGS_MODULE: specifies which settings module to use
• JAZZ_API_KEY: retrieves jobs from jazzhr api at /api/jobs/
• NPM_CONFIG_PRODUCTION: set to false so npm also install devDependencies docs
• RECAPTCHA_SECRET_KEY: https://www.google.com/recaptcha/ keys for use in subscription forms and in /api/subscribe
• RECAPTCHA_SITE_KEY: same as above
• WEB_CONCURRENCY: read the docs

Handy Commands

• heroku run "./manage.py shell" --app newamericadotorg: open a shell in production environment
• heroku run "./manage.py sync_campaign_monitor": should probably just add this to heroku scheduler and run every night
• heroku pg:pull PRODUCTION_DATABASE_NAME mylocaldb --app newamericadotorg: dump production database to local database read the docs
• heroku pg:copy PRODUCTION_DATABASE_URL STAGING_DATABASE_NAME --app na-staging: copies data from production database to staging. read the docs. Use with caution.
  • After copying the database from production to staging, you may need to copy over images to the s3 staging bucket: aws s3 sync s3://newamericadotorg/images s3://na-staging/images
• to export lists of wagtail users:
  • run heroku pg:psql --app YOUR-APP-HERE
  • in the postgres shell, run \copy public.auth_user (email, is_active) TO '/LOCAL/FILE/LOCATION.csv' CSV HEADER;