Replication: When to Use and When Not to Use BBR

This repository contains the artifacts for

Soumyadeep Datta and Fraida Fund. 2023. Replication: When to Use and When Not to Use BBR. In Proceedings of the 2023 ACM Internet Measurement Conference (IMC 23), October 2426, 2023, Montreal, QC, Canada. ACM, New York, NY, USA. https://doi.org/10.1145/3618257.3624837

which replicates part of the results in

Yi Cao, Arpit Jain, Kriti Sharma, Aruna Balasubramanian, and Anshul Gandhi. 2019. When to use and when not to use BBR: An empirical analysis and evaluation study. In Proceedings of the Internet Measurement Conference (IMC '19). Association for Computing Machinery, New York, NY, USA, 130136. https://doi.org/10.1145/3355369.3355579

Reproducing the figures using our experiment data

You can use our experiment data directly to generate the figures in our replication paper. Use the materials inside the data directory, or:

Recreating our experiment and generating data on FABRIC

You can use the materials in this repository to recreate our experiment on the FABRIC testbed and generate new data yourself. Assuming you already have a FABRIC account, have set it up, and are part of a project (if not, see Hello, FABRIC), you can open the FABRIC Jupyter environment, and in a new terminal, run:

git clone https://github.com/sdatta97/imcbbrrepro

then open the fabric-run.ipynb notebook inside the fabric-notebooks directory. Follow the instructions in that notebook to configure your experiment, then run it.

Recreating our experiment and generating data on CloudLab

You can use the materials in this repository to recreate our experiment on the CloudLab testbed and generate new data yourself. Assuming you already have a CloudLab account, have set it up, and are part of a project (if not, see Hello, CloudLab), you can instantiate an experiment using the link:

https://www.cloudlab.us/p/nyunetworks/bbr-when-to-use

Select the Ubuntu 18.04 disk image to reproduce our replication, or the Ubuntu 20.04 disk image to reproduce our extension to BBRv2. Then, continue to reserve resources and start your experiment on any cluster.

Once your resources are reserved and active, you can SSH into each of the nodes and run experiments using the instructions provided inside the cloudlab-materials directory.

This material is based upon work supported by the National Science Foundation under Grant No. 2226408. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.