FCProto

This repository contains a prototype developed for the Fog Computing lecture. The project demonstrates the key principles and practical applications of fog computing, including distributed data processing and edge device coordination.

Cmd

The directory contains the main entry points for the cloud and edge applications.

Deployment

Configuration files for Docker and Terraform to deploy the application.

Docker

Start the cloud environment with the following command:

docker compose up cloud

Start the edge application with the following command:

docker compose up edge

The docker image repository can be found here.

Terraform

To provision the cloud infrastructure with Terraform, navigate to the terraform directory and run the following commands:

terraform init
terraform validate
terraform plan
terraform apply

Ansible & Kubernetes

We originally tried to deploy the application with microk8s and ansible to make the usage scenario as realistic as possible. Unfortunately, we did not manage to get microk8s to use the cloud node completely. It was only classified as ready, but no pods were released. It only worked locally to run the fog computing deployment with node affinity. We thought it was a pity to throw away the big effort and therefore decided to keep the code in the repository.

Docs

The docs directory contain documentation for the project like the demonstration video and the project description report.

show-case-recording.mp4

Open the prototyping assignment report here.

Internal

Cloud

The cloud directory contains logic that is only executed on the cloud side. This includes env configurations, database logic with entities and logic to start up server.

Edge

The edge directory contains logic that is only executed on the edge side. This includes env configurations and logic to start up the edge.

Shared

The shared directory contains logic that is shared between the cloud and edge side. This includes the message model, queue logic, connection logic and utility functions.

Pkg

The pkg directory contains all sensor logic developed for the project.

Sensor

Sensor interval can be adjusted in the sensor client. The sensor logic is responsible for generating data and run on the edge application. There are three different sensors implemented:

• VirtualSensor: Generates random temperature & humidity data.
• MemorySensor: Generates memory usage data.
• CpuSensor: Generates CPU usage data.