DNS Privacy with Speed? Evaluating DNS over QUIC and its Impact on Web Performance

Mike Kosek¹ | Luca Schumann¹ | Robin Marx² | Trinh Viet Doan¹ | Vaibhav Bajpai³

¹ Technical University of Munich | ² KU Leuven | ³ CISPA Helmholtz Center for Information Security

IMC 2022, October 25–27, 2022.

[Paper Arxiv →] https://arxiv.org/abs/2305.00790

[Paper ACM DL→] https://doi.org/10.1145/3517745.3561445

---

Tools

The following tools were enhanced for our paper;

1. DNSPerf: Performance measurement library for DoQ, DoUDP, DoTCP, DoT, and DoH

• Repository: https://github.com/mgranderath/dnsperf
• Pull Request adding support for QUIC Address Validation as well as TLS 1.3 Session Resumption and 0-RTT for DoQ, DoT, and DoH: mgranderath/dnsperf#3

2. DNS Measurements: Performance measurement tool for DoQ, DoUDP, DoTCP, DoT, and DoH

• Repository: https://github.com/mgranderath/dns-measurements
• Pull Request adding support for QUIC Address Validation as well as TLS 1.3 Session Resumption and 0-RTT for DoQ, DoT, and DoH: mgranderath/dns-measurements#2

3. DNS Proxy: Simple DNS Proxy Server supporting DoQ, DoUDP, DoTCP, DoT, and DoH.

• Repository: https://github.com/AdguardTeam/dnsproxy
• Fork used in the paper: https://github.com/justus237/dnsproxy
• Pull request adding support for QUIC Address Validation as well as TLS 1.3 Session Resumption and 0-RTT for DoQ, DoT, and DoH: AdguardTeam/dnsproxy#268
• Pull request DoT connection reuse: AdguardTeam/dnsproxy#269

---

Reproducibility

In order to enable the reproduction of our findings, we make the developed tools, the raw data of our measurements, and the analysis scripts publicly available. Please note, that our analysis scripts use ip-api (https://ip-api.com) for IP-to-Geolocation mapping. Due to changes in IP ownership, the data derived from the API might change over time. For our paper, we queried the APIs on the 16.05.2022.

0. Repository Overview

• The files single.query.response.times.ipynb, single.query.sizes.py and web.performance.ipynb are the analysis scripts for the , single query sizes, single query response time and web performance measurements
• The folders single.query.response.times.zip, single.query.sizes.zip and web.performance.zip contain the dataset as well as supplementary files for the single query response time, single query sizes and web performance measurements
• The file single.query.sizes.processed.ods are the aggregated results of the single.query.sizes.py script

1. Preparations

• Clone this repository to a machine running (Jupyter Notebook or JupyterLab) and python3
• Extract single.query.response.times.zip to the subdirectory single.query.response.times/
• Extract single.query.sizes.zip to the subdirectory single.query.sizes/
• Extract web.performance.zip to the subdirectory web.performance/

3. Single Query Response Times

• Run the Jupyter Notebook single.query.response.times.ipynb

4. Single Query Sizes

• Run the python script single.query.sizes.py

5. Web Performance

• Run the Jupyter Notebook web.performance.ipynb

---

Contact

Please feel welcome to contact the authors for further details.

• Mike Kosek ([email protected]) (corresponding author)
• Luca Schumann
• Robin Marx ([email protected])
• Trinh Viet Doan ([email protected])
• Vaibhav Bajpai ([email protected])