Continuum

Continuum is a deployment and benchmarking framework for the edge continuum. It automates setting up and configuring emulated cloud, edge, and endpoint/IoT hardware and networks, manages installation of software inside the emulated environment, and can perform application- and system-level benchmarks.

CONTINUUM IS UNDER ACTIVE DEVELOPMENT

How it works

Continuum has the following architecture:

The execution flow consists of three phases, each having a configuration and execution step. The phases are infrastructure deployment, software installation, and benchmarking. Each phase can be skipped, i.e. the framework can be used for infrastructure deployment only if so desired.

1. Infrastructure configuration: Libvirt configuration files are created based on the user's preferences.
2. Infrastructure execution: The configuration files are executed, creating QEMU/KVM virtual machines connected through network bridges.
3. Software configuration: Based on the configured infrastructure, Ansible is configured for software installation.
4. Software execution: Ansible playbooks are executed, installing operating services and resource management software on each machine. This includes setting up resource management clusters such as Kubernetes.
5. Benchmark configuration The benchmark is configured and prepared based on the user's preferences.
6. Benchmark execution: Applications in docker containers are executed on resource management software running on the emulated infrastructure (Kubernetes, KubeEdge, etc.). Meanwhile, application- and system-level metrics are captured and finally processed and presented to the user.

Installation

The framework supports execution on multiple physical machines through a network bridge. Continuum has been tested on Ubuntu 20.04. Recommended software versions are noted below.

1. Install Docker on each physical machine (tested with v20.10.8)
2. Install KVM, QEMU and LibVirt on each physical machine (QEMU v4.2.1 or v6.1.0).
   
   • For execution on a single physical machine, the virtual network bridge virbr0 will be used. Most often, this will be automatically configured when installing KVM/QEMU/LibVirt.
   • For execution on multiple physical devices, a network bridge br0 is required. This requires extra configuration, see docs/NETWORK.md.
   • Run export LIBVIRT_DEFAULT_URI="qemu:///system" and/or add to your .bashrc or similar.
3. Install Python 3 (v3.8.10).
   • Install the following packages: numpy pandas. For reproducing experiments from our work, matplotlib is required as well.
4. Install Ansible on each physical machine (tested with v2.9.6).
   • Add the following to ansible.cfg (typically in /etc/ansible/ansible.cfg):
     • Under [ssh_connection], add retries = 5
     • Under [defaults], add callback_whitelist = profile_tasks

Please read docs/ISSUES.md when encountering issues during the installation or usage of the framework.

Usage

Please follow these steps to run the framework. The framework needs to be executed on a single physical machine, even if you want to make use of multiple physical machines.

1. Clone this repository: git clone ...
2. Change directories to the edge resource manager you want to benchmark (example: cd KubeEdge)
3. Run the main python script to start the benchmark: python3 main.py. For more information on how to use the script, see python3 main.py -h.