Suppose a computer user is running multiple Tor clients: a system-wide tor instance, Tor Browser, Tor Messenger, Ricochet, etc.

Given that all of these clients independently select entry guard nodes and that this likely increases the risk of selecting an adversary's node to be a guard, is it better to:-

  1. have each client use the same tor instance, say, the system tor client?
  2. configure each client with hard-coded entry guards, e.g. let one of the clients select some guards and configure the rest to match (and periodically reconfigure as appropriate)?
  3. accept the increased risk and do nothing?

2 Answers 2


TL;DR The answer will depend on what you consider important, and what your Tor use-case is. (Similar to the hand-waving over threat models mentioned in another answer.)

However, in general, the recommendations would be to:

  • Use fewer guards;
  • Keep the guards for longer.

These are described in Part 3 of a post on the offical Tor blog, which references 3 different research papers:

A very quick synopsis.

...I think changing the guard rotation period to a year or more is probably much wiser...

While this last point doesn't explicitly mention hard-coding a single guard - as you suggested in your question - it doesn't recommend against it.


I propose you the next model :

  • use multiple ISP's, run a pair ISP+tor client, ban them from using each other
  • use a "gateway" client that uses pair clients as the points it uses to go to Tor network
  • configure your tor clients to use a gateway client as their Tor exit

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .