Philosophy

Happy comes from the mindset of making the secure thing easy.

At CZI, we have lots of little and medium sized applications. These apps are generally simple HTTP web applications. Many times, developers don’t have the capacity or time to wade through the 1000s of ways to deploy or stand up their application. Happy is now the de-facto way of deploying applications in minutes at CZI. For folks with more advanced use cases, it is a great starting point to iterate on, while being robust and flexible enough to add advanced infrastructure or features. For folks with simple us cases, it just works.

Technologies

It uses the following technologies:

• Containerized applications
• Docker compose
• EKS or ECS clusters
• Terraform
• Happy CLI tool
• Github Actions

Features

It gives your application out-of-the-box:

• Auditing and logging
• End-to-end encryption
• Incident response
• Load balancing and autoscaling
• Service discovery with other happy stacks
• Development environments
• Shell access to stacks
• Automatic deployment workflows

Composition vs Prescription

Happy is more of a composition of tools rather than a rigid prescription used to deploy an application. We found that each application is different in small ways and reflects the culture and style of each team. To try and accomidate all the styles and teams would make the tool useless. And to force teams to fit a single style would stiffle new ideas.

We settled on a composition of tools that work together to provide an easy deployment path. If you are more familiar with kubectl, there is nothing stopping you from exploring your cluster that way. If you like to look at your deployment through AWS console, go ahead. If you hate using docker-compose for local development, don’t use it.

For those less familiar with infrastructure or find yourself saying “I just want to deploy my app and see it”, then happy provides a defined path and set of tools to do that.