FSA Deploy Juke

What’s that? It’s time for another Deployment tutorial! The last three (trip planner, twitter.js and wikistack) went pretty well but here we have another challenge - seed media files on top of simply seeding a PostgreSQL database!

So the core instructions are sited at the wikistack tutorial. Do not push to heroku until i tell you to. Here I will focus on the incremental difference.

  1. to configure process.env.DATABASE_URL, go to server/db/db.js and add it in there alongside require(path.join(__dirname, '../env')).DATABASE_URI
  2. old pg dependency is gone
  3. remember to create a table in SQL. this time, call it juke. refer to the wikistack tutorial for instructions
  4. ok so it gets tricky from here! We aren’t going to use webpack to rebuild the app on heroku so we need public/bundle.js to be served. delete that and public/bundle.map.js from the .gitignore so you will be able to push it to heroku. you may want to do this in a different branch (git checkout -b herokubranch) so you still have the right gitignore for github, but i didn’t bother. (nothing secret in public anyway).
  5. remove the postinstall process from package.json. you’ll seed your database manually…
  6. Time to push to heroku! git add . && git commit -m "heroku deploy". MAKE SURE BUNDLE.JS is ADDED!! and now when you push here the difference is you are pushing a specific BRANCH instead of master. However, heroku only runs master. how to do that? git push heroku <YOURBRANCHNAME>:master.
  7. ok hopefully that went well. to seed the db, heroku run node ./bin/seed
  8. heroku open!

contact me @swyx on twitter if you run into any issues. My app is deployed here https://swyx-juke.herokuapp.com/#/