Skip to content

Power Consumption Analysis of Tor Bridges Stalled Due to Network Congestion, Seeking Guidance on One-Hop Relay Setup

Description:

We are conducting an experiment to analyze the factors affecting power consumption on a Tor bridge. Our methodology involves connecting multiple clients to a single bridge and measuring power usage while varying several parameters, including:

  • Number of concurrent clients: Testing how the client load impacts power draw.
  • Connection duration: Analyzing the effect of long-running versus short-lived connections.
  • Type of data transfer: Differentiating between light browsing, bulk data transfers, and streaming to see how each traffic pattern influences power consumption.

The Problem:

Our experiment has been significantly hampered by the ongoing network congestion across the Tor network. We believe the current events in Iran and Israel have led to a surge in Tor usage, causing bootstrapping to take a long duration. This congestion is introducing too many confounding variables into our measurements, making it impossible to isolate the factors we intend to study. The reliability of our power consumption data is compromised as we cannot distinguish between power draw from our experimental variables and power draw from handling network instability.

Proposed Solution:

To circumvent the current network-wide congestion and obtain a cleaner dataset, we propose to modify our experiment to use a one-hop relay. By creating a direct, single-hop circuit from our clients to our bridge (acting as a relay), we can bypass the public, congested parts of the Tor network. This will allow us to create a controlled environment to measure the impact of our variables more accurately.

Question:

We are looking for guidance on how to configure our clients and our bridge to force a one-hop relay for our experiment. What are the necessary torrc configurations or command-line flags to achieve this? Are there any best practices or potential pitfalls we should be aware of when setting up a private, one-hop Tor circuit for testing purposes?

Any assistance or pointers to relevant documentation would be greatly appreciated!