Setup Guide

This guide explains the necessary steps to setup a new FlockLab testbed.
A FlockLab testbed consists of a server part and an observer part. The observers interact with the devices under test (targets) and locally store the test data. The server provides a user interface and is responsible for starting and stopping tests as well as fetching the test data from the observers.

Server

The FlockLab server must provide the following three components:

• test management
• database
• webpage

These components can either be on the same machine or distributed on separate servers.

Test management

• Copy the files to your test management server.
• Adjust the config file and copy it to the server.
• Adjust the path to the config file in flocklab.py.
• Install crontab to periodically call the scheduling and cleanup scripts.
• Install all required python packages.
• Make sure the user that runs the test management scripts has SSH access to all your FlockLab observers via public key authentication.

Database

• Create a MySQL-compatible database flocklab on your DB server.
• Use the provided template to create all necessary tables.
• Create the user flocklab_testmng and grant it access to the database flocklab.
• Start to populate the database tables: tbl_serv_observer, tbl_serv_architectures, tbl_serv_platforms, tbl_serv_tg_adapt_types and tbl_serv_tg_adapt_list.
• Make sure the database access credentials in flocklab_config.ini on the test management server are correct.

Webpage

• Copy the webpage files to your webserver.
• Adjust the path to the flocklab_config.ini file in include/config.php.
• Adjust the XML validation template xml/flocklab.xsd.
• Adjust the floor plan (pics/flocklab_floormap.png) and node positions (scripts/flocklab-observer-positions.js).

Note: you can use the update server script to copy the files to your test management and webserver (don't forget to adjust the HOST variable).

Observer

Hardware setup and calibration

A guide on how to prepare the observer hardware in available on request.

BeagleBone Green setup

• Download the latest Debian flasher image (console version is sufficient) from eLinux (monthly builds) or beagleboard.org (releases). Use e.g. Balena Etcher to write the image to a microSD card.
• Insert the microSD card and press and hold the user button while powering up the BeagleBone until all 4 LEDs are on. The Debian image will then be flashed onto the eMMC (takes ~10 minutes). After completion, the BeagleBone will shut down (all LEDs off).
• Replace the flasher SD card with an empty one (formatted to ext4).
• Make sure the network connectivity to the BeagleBone works.
• Clone the repository to your computer and run the basic setup script: cd observer/debian && ./setup_system.sh [beaglebone_address]
• Adjust the parameters in the following scripts:
  • HOSTNAME in the ping watchdog script
  • HOSTPREFIX and OBSIDS in the observer update script
• Run the main setup script to install all required packages, tools and scripts on the BeagleBone: ./setup_observer.sh [beaglebone_address]
• Pair the BeagleBone with a FlockLab 2 cape and install the correct calibration files.