IoT with Django: From hackathon to production by Anna Schneider
I was recently interviewed for an episode of the Talk Python to Me podcast. The theme of the episode was how we at Opbeat ship, run, and monitor Python applications.
Each larger release of Django includes a multitude of features, fixes and the like. In the following we will take a look at some of the more prominent features of the Django 1.9 release.
Andrew Godwin (@andrewgodwin), Django Core Contributor & Eventbrite Engineer, talks about building a Heroku for Python, migrating from MySQL to PostgreSQL without downtime, and what Kickstarter can do for the Django community.
Jacob Kaplan-Moss (@jacobian), Django contributor and Director of Security at Heroku, talks with us about managing continuous deployment at Heroku, how they use intrusion detection systems to keep things safe in house, and shares tips on ways that smaller teams can improve their security processes.
Our main web app is written in Django and run in uWSGI. We use AWS Elastic Load Balancer to distribute requests between our webservers and we use Fabric to make sure new code get distributed to webservers and signal uWSGI to reload the app. Django is notoriously lazy at loading apps, so as part of the deployment process our Fabric script automatically shoots off a few requests to the server to get it warmed up.
At Opbeat, we use Django extensively. With the new
ALLOWED_HOSTS setting, you need to make sure
ALLOWED_HOSTS cover all the domains you will be serving from your Django application.