[tor-bugs] #9386 [Tor]: Prop 221: Use CREATE_FAST less, or not at all

[Tor Bug Tracker & Wiki]

#9386: Prop 221: Use CREATE_FAST less, or not at all
-----------------------------+---------------------------------------------
     Reporter:  nickm        |      Owner:
         Type:  enhancement  |     Status:  needs_review
     Priority:  normal       |  Milestone:  Tor: 0.2.5.x-final
    Component:  Tor          |    Version:
   Resolution:               |   Keywords:  tor-client prop221 024-backport
Actual Points:               |  Parent ID:
       Points:               |
-----------------------------+---------------------------------------------

Comment (by nickm):

 Replying to [comment:6 arma]:
 > Looks better yes. I wonder if there are other edge cases.

 There likely are.  Can you think of any likely way to find them better
 than "read the code hard" or "merge and see what happens"?

 > As for the attack you describe, yes. I think that used to be possible
 too by answering with a created-fast and then just dropping any extend or
 begindir cells?

 That's also true.

 > I guess another option here is what you were heading toward originally:
 we dump the guard if it ever can't process one of our create requests.
 That seems a poor plan though, especially in light of the current circuit
 create load?

 Probably so.

 Shall we merge, or think more?

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/9386#comment:7>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online