How can I be sure, that the entry node, middle node and the exit node are not operated by the same operator?
Of course there is a small chance that this happens. But can the operator of the entry node control which node is going to be the middle node? So that the middle node can be a node also operated by him.

If so the attacker (which controls the entry node, middle node and exit node) could then strip away all encryption layers and see my traffic, which would make tor useless.

Wouldn't it be more secure if there were no entry guards and my client would behave as a middle node even if it's the client? So the first node can not know if I'm a middle node or a client. I can't see why this is not the case in tor.

  • 1
    The client chooses all of the relays in a circuit, not the guard relay. Tor tries its best to limit the probability of two relays in a single circuit being run by the same operator, but the answer would be that you can never be 100% sure. – Steve Jan 10 at 22:54
  • So the client tells the entry node: extend my circuit to this ip. Can the entry node now create an own middle node (in their own internal Network) and redirect traffic to the newly created node? – redstoner2014 Jan 11 at 18:28
  • No because the client downloads the consensus from the directory authorities, which includes an “identity key” for each relay. I’m not sure the exact process, but the client will authenticate the new relay each time it extends the circuit. – Steve Jan 11 at 20:04
  • Okay cool, so unless the directory authorities are not controlled by an attacker (or the attacker is redirecting the request to his own directory) the chance that the entry node and the middle node are controlled by the attacker are pretty low. – redstoner2014 Jan 13 at 2:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.