This is a Flask-based frontend framework that acts as abstraction layer for a digital video publisher's OVP, like a Brightcove or Kaltura service. Typically, enterprise users of video hosting and streaming solutions need to provide very limited access to a large population of users. However, desipite user profile permissions restrictions, most OVP's cannot completely insulate valuable video assets, which in most scenarios are tracked, monetized, and assigned metadata, from the end user.

This abstraction layer uses Gunicorn and PostgreSQL, and is configured for Heroku deployment with easy integration into AWS S3.

Getting Started:

To start gunicorn server locally you need to swap out the configuration in wsgi.py to "production" instead of "production_heroku".

Then run: $ gunicorn app.wsgi:application

For errorhandler() to work, have to set Debug=False in the config files. Also, for Gunicorn, drop the '--log-level debug' from $ gunicorn app.wsgi:application --log-level debug command. The full command would be used locally in dev environment. In Flask, if you set Debug=True, it re-routes all the errors to the console, so the builtin errorhandler that's used to display custom error pages in routes.py via render_template doesn't work.

(Leaving run.py file in case you want to use Flask server instead of Gunicorn)

For Heroku deployment, just make sure the wsgi.py is pointing to "production_heroku"

Some userful Heroku commands:

Logging:

heroku logs --tail

Access the shell of your one-off dyno:

heroku run bash

Also, if you want to work on the same repository that is for the same Heroku app, you need to connect the repository copies on your other machines to the same app via:

$ heroku git:remote -a yourappname

The other way to get the actual repository that's used for the app is to git clone it from the Heroku dashboard, but that won't actually be connected to your github repository version unless you create a new github repository, push it, etc.

For local dev, sometimes port gets stuck -- to find offending port: $ lsof -n -i4TCP:8000 | grep LISTEN

AWS S3:

Basically we're using a public-access bucket on AWS S3 to upload video assets and be able generate/grab publically available file URLs so that they can be passed to Brightcove's Dynamic Ingest API (although this is useful for any other OVP with a similar ingest option).

To configure your S3 bucket, here's an easy way that works as of this writing:

1. Sign into Amazon Web Services and go to S3 Management Console

2. Click the Properties button for the desired bucket

3. Go to Permissions tab and hit Add Bucket Policy link (or Edit Bucket Policy if you've already added one previously).

4. Copy and paste the following into the Bucket Policy Editor popup. Replace "YOUR-BUCKET-NAME" with your own bucket's name:

{"Version": "2008-10-17", "Statement": [{"Sid": "AllowPublicRead", "Effect": "Allow", "Principal": { "AWS": "*" }, "Action": "s3:GetObject", "Resource": "arn:aws:s3:::YOUR-BUCKET-NAME/**" }]}

It doesn't seem like this should work, and you should use the Policy Generator ( http://awspolicygen.s3.amazonaws.com/policygen.html ), but the above really does work.