We teamed up with our friends over at Gestalt IT to present (and record) a more thorough demo of our solution in front of an esteemed list of Tech Delegates.
The first section is an introduction to Avassa: why we do what we do and how we do it. It provides a short overview of our tech-stack's moving parts, their respective role in the system and how they interact. If you haven't heard about us yet, start here before moving into more specific topics.
Explaining platform technology can be challenging as it is harder to make it "come alive" when compared to explaining systems that are single purpose. You run the risk of being too vague and non-descriptive, leaving the responsibility on the viewer to connect the dots. To alleviate this, we put together a demo universe built on the premise that we control the IT-environment in a chain of movie theaters. In this first section we recruit the team, walk through the conditions we have, introduce a partner and set up a small list of requirements that we are going to try and meet.
Then we build an edge cloud consisting of more than a hundred sites.
Now that we have a cloud, it's time to deploy our first application. In this video, we focus on our top priority: deploying new popcorn-controller software to ensure timely delivery of quality popcorn to our customers. It's a very simple application we deploy to all our more than hundred sites in one single operation.
We then move to a more involved application to manage our theater-rooms. This is a multi-container application that makes use of our distributed, secure key-value store for managing application secrets. This also introduces versioning of both the application and the constituent containers (using image tags). To make it more real, we deploy and then upgrade this application using GitLab where we store the application and deployment specifications, then push them on-change using GitLab Runners.
The next step is to monitor and observe the applications.
Now that we have a distributed edge cloud with more than a hundred sites and two applications running on it, it's time to focus on how to monitor and observe them.
In this video, we set up a simple exporter service to feed metrics and observability data related to our application running across the sites into AWS CloudWatch.
We wrap things up by introducing application tenancy into the system. This allows the administrator to assign site-level resources to specific tenants while keeping a deep resource isolation between them. No data is shared and is encrypted per-tenant both at rest in the edge sites and in-flight. We also show how tenants get the same tools experience as the administrator to interact with the system.