Introduction to By Michael Mottola

What is Heroku?

Heroku platform builds your application from your source code and dependency descriptions.
Executes a command to run your application (ex: starting a server)
How you define your application and dependencies varies:
- Ruby uses Gemfile
- Python uses requirements.txt
- Node.js uses package.json
- Java uses pcm.xml

If using an established framework, Heroku figures it how to run your application:
- Ruby is rails server
- Djano is python <app>/manage.py runserver
- Node.js uses main field in package.json
Else you must explicitly declare how your app is executed:
- Create text file called Procfile
```
web: bundle exec rails server -p $PORT
```

Lightweight Linux containers that runs a single user-specified command, called a Dyno
Single dyno can serve thousands of requests per second
Types of Dynos:
- Web Dynos: only receive HTTP traffic from Heroku's routers
- Worker Dynos: used for background jobs, queueing and timed jobs
- One-off Dynos: temporary, runs detached or from terminal
Under the hood, Heroku is using Amazon Web Services; EC2 instances

Inbound requests are received by a load balancer and passed to a set of routers
Routers determine location of your application's web dynos and forward the HTTP request

Heroku Command Line Interface (CLI) for managing and scaling applications and add-ons.

https://toolbelt.heroku.com/

$ heroku login
Enter your Heroku credentials.
Email: michael@example.com
Password (typing will be hidden):
Authentication successful.

$ heroku create my-app
Creating my-app... done, stack is cedar-14
http://my-app.herokuapp.com/ | git@heroku.com:my-app.git

Deploy command:

git push heroku master

Scaling Horizontally: add more dynos
- more web dynos means more concurrent HTTP requests
Scaling Vertically: using bigger dynos (increase RAM)

Scaling with command line:

heroku ps:scale web=2