The repository is used to reproduce the evaluation from
End-To-End Timing Analysis and Optimization of Multi-Executor ROS 2 Systems
for RTAS 2024.
To replicate the experiment in the paper, please follow the instructions to either set up the VM or set up the packages. Afterwards, install Gurobi, and follow the steps on how to run the experiments.
This document is organized as follows:
-
Please download the zip file, which contains the virtual disk and the machine description.
-
Install VirtualBox https://www.virtualbox.org/.
-
Import the .ova file in Virtualbox and start the virtual machine.
-
The credentials are: ros2end2end/rtas2024 (user/password).
-
Start a terminal (Press CTRL+Shift+T) in the virtual machine and run the following command:
./initialize.sh -
Afterwards, you can use the README in the folder ros2-end-to-end-distributed to install Gurobi (see Gurobi Installation) and run the experiments.
-
Start a terminal (Press CTRL+Shift+T) and run the following commands:
sudo snap install code --classic sudo apt -y install git python3-pip pip install gurobipy tabulate git clone https://github.com/HarunTeper/ros2-end-to-end-distributed.git -
Afterwards, you can use the README in the folder ros2-end-to-end-distributed to install Gurobi (see Gurobi Installation) and run the experiments.
These instructions explain how to install Gurobi on the virtual machine. They are based on the state of the Gurobi webpage on 15.02.2024. For this evaluation, Gurobi 10.0.3 is required.
-
Go the the following page: https://www.gurobi.com/.
-
Register a new account if you do not already have one.
-
Log in to your account.
-
You should now be on your user home page https://portal.gurobi.com/iam/home.
-
Request a license. For example, go to Request a free academic license (https://www.gurobi.com/academia/academic-program-and-licenses/).
-
Choose an Academic Named-User License and click on Learn More.
-
Follow the steps described on the website (https://www.gurobi.com/features/academic-named-user-license/).
-
The following website provides a complete guide for the setup (https://support.gurobi.com/hc/en-us/articles/4534161999889-How-do-I-install-Gurobi-Optimizer).
-
Click the Download Gurobi Optimizer button and accept the EULA.
-
Download Gurobi 10.0.3, which is an older version and can be found below the current version of Gurobi 11.0.0.
-
Download Gurobi for your respective operating system. For the VM, it is the gurobi10.0.3_linux64.tar.gz file.
-
Extract the gurobi folder that is in your Downloads folder into your home directory.
-
Set the paths of GUROBI_HOME, PATH, and LD_LIBRARY_PATH. A detailed description can be found here https://support.gurobi.com/hc/en-us/articles/13443862111761-How-do-I-set-system-environment-variables-for-Gurobi. Please note that this description may be based on an older version, and you may need to change the folder names according to the downloaded files.
-
For the VM, the following commands works for our configuration of the .bashrc:
export GUROBI_HOME="/home/osboxes/gurobi1003/linux64" export PATH="${PATH}:${GUROBI_HOME}/bin" export LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:${GUROBI_HOME}/lib"
-
-
Revisit the page https://portal.gurobi.com/iam/home.
-
Click on Licenses.
-
Click on Request next to Licenses.
-
Request a Named-User Academic license.
-
Agree to the End User License Agreement.
-
Click on Confirm Request.
-
Click to copy the command grbgetkey including the key that is displayed.
-
Open a terminal, copy and enter the command into the terminal. Accept any prompts that come up during this setup.
-
To test the gurobi installation, open a terminal and run the following command:
gurobi.shIf this command lists the configuration of your installation and opens a gurobi terminal, the installation is successful, and you may proceed to run the experiment. You may close the Gurobi terminal again.
The experiment involves running the optimization of the evaluation. Specifically, we provide a script that calculates the upper bound values (UB) of our system. The reference values can be found in the rightmost column in the table of our evaluation, and are as follows:
| Configuration | UB |
|---|---|
| DDS | 696.045 |
| Timer Periods | 668.146 |
| Policy | 665.084 |
| Assignment | 832.429 |
| Fix-Async | 416.177 |
| Fix-Sync | 493.984 |
| Fix-Assign | 419.653 |
| Baseline | 835.837 |
During the artifact evaluation process, we also had to update our optimization parameters. Below, you can find the values that were output by an earlier commit in this repository:
| Configuration | UB |
|---|---|
| DDS | 0.671045 |
| Timer Periods | 0.668146 |
| Policy | 0.665084 |
| Assignment | 1.13843 |
| Fix-Async | 0.476134 |
| Fix-Sync | 0.512733 |
| Fix-Assign | 0.419653 |
| Baseline | 0.835837 |
Open a terminal and move to the folder of this repository (ros2-end-to-end-distributed)
Run the following commands:
git pull
python3 main.py
After the script finished, the results are shown in the terminal.
From our experience, running the optimization requires less than ten seconds.
ros2-end-to-end-distributed
├── system_configurations
│ └── indy.yaml
├── buffer.py
├── callback.py
├── executor.py
├── main.py
├── node.py
├── optimization.py
├── publisher.py
├── README.md
├── subscription.py
├── system.py
└── timer.py
The file optimization.py corresponds to the optimization implementation and it includes comments that reference to the corresponding lemmas of the paper in Lines 321-374
As a reference, we utilize a machine running Ubuntu 22.04, with an AMD Ryzen 5900 12-Core Processor (12 Cores, 24 Threads) with 3,7GHz and 32GB RAM.
For the virtual machine, we enabled 4 cores and 4096 MB of RAM.
- Harun Teper
- Tobias Betz
- Mario Günzel
- Dominic Ebner
- Georg von der Brüggen
- Johannes Betz
- Jian-Jia Chen
This work has been supported by the Federal Ministry of Education and Research (BMBF) in the course of the project 6GEM under the funding reference 16KISK038.