fogify-demo

Demo: Emulating Geo-Distributed Fog Services

This demo presents an end-to-end system that combines Fogify, our interactive fog emulator, with Jupyter, a web-based interactive tool for data analysis. On the one hand, Fogify helps developers by enabling the deployment of emulated fog realms locally or on cloud infrastructure while easing the description of fog topologies by extending the Docker Compose to support the "fogified" model specification. With the "fogified" model, Fogify allocates resources as separated containerized processes, and establishes network connections between them. At running time, Fogify allows developers to inject ad-hoc faults, entity downtime, perform scaling actions, adjust the workload, network changes and restrict data movement.

Requirements

Before starting, we have to install docker, docker-compose and docker swarm on the infrastructure. For more information, we suggest the official documentation.

Furthermore, the system executes some low-level commands in order to apply specific characteristics to the network. For this reason, on each cluster of the swarm cluster, we should install the traffic control tool (tc-tool). On Debian-based destributions, tc-tool comes bundled with iproute, so in order to isntall it you have to run:

apt-get install iproute

Stack Instantiation

The Fogify and Jupyter stack is provided as dockerized services described via docker-compose file. So we should run the following command at the swarm master node.

sudo docker-compose -p fogemulator up

In order to open the web interface of Jupyter, you have to find the output of the ui service where the system outputs the url and the token

For instance, in previous output we can open the Jupyter interface with the following url:

http://127.0.0.1:8888/?token=1cd2e914cd03e76d551a666cb0a8dcdb6361bc29ddf54eed

If we have more nodes in the cluster, we have to execute the following command on every node.

sudo docker-compose -p fogemulator up agent cadvisor

For both modes, users should provide a set of parameters to compose file (written in .env file), specificaly:

MANAGER_NAME=.... #The name of the swarm master node
MANAGER_IP=192.168.1.1 #The IP of the swarm master node
HOST_IP=192.168.1.1 #The IP of the host

Preparation of taxis use-case

As a simple testing app, we implemented a simple IoT service that is driven by real-world data (New York Taxi) to showcase a scenario of a taxi-cab company that collects and analyzes location-based data from its fleet. The codebase of the application is under the application folder and can be easily build by running the following command:

./application/build-image.sh

When the containers are built, user has to create a folder named /home/ubuntu/data and should place the yellow_tripdata_2018-01.csv of New York taxi trip dataset. Finally, the user can utilize the demo_files/docker-compose.yaml file as input of the Fogify.