This is for all my tips, snippets, and gotchas
{
"scripts": {
"dokku": {
"predeploy": "sputnik --name spacy --repository-url http://index.spacy.io install en==1.1.0"
}
}
}
- Use Procfile for processes
- Use requirements.txt for python dependencies
- Use nltk.txt for NLTK downloads
- Use apt-packages for installs via apt-get
pg_dump --host=aa1xsbm3n9zfguj.c9hgttwrfmsn.us-west-1.rds.amazonaws.com --port=5432 --username=trivianote --password --dbname=ebdb --no-owner --no-privileges --no-acl -Fc > backup-production-2017-12-22.dump
- -Fc is for custom format, which allows the .dump file type
pg_restore --verbose --exit-on-error --single-transaction --host=localhost --port=5432 --username=josh --dbname=test --no-owner ./backup-production-2017-12-22.dump
web: bundle exec puma -C config/puma.rb web2: gunicorn --config config/gunicorn.py application:app
- Keep config files like gunicorn.py out of the root directory, because it will cause errors when running commands like "gunicorn app:app"