[tor-bugs] #1826 [Tor Relay]: Sending packets after RELAY_END

[Tor Bug Tracker & Wiki]

#1826: Sending packets after RELAY_END
-----------------------+----------------------------------------------------
 Reporter:  mwenge     |       Owner:                    
     Type:  defect     |      Status:  new               
 Priority:  normal     |   Milestone:  Tor: 0.2.2.x-final
Component:  Tor Relay  |     Version:  Tor: unspecified  
 Keywords:             |      Parent:                    
-----------------------+----------------------------------------------------

Comment(by mwenge):

 Replying to [comment:10 mwenge]:
 >
 > To close this out I will follow Nick's suggestion and perform this test
 with my Tor acting as both client and exit and ensure the cell queue isn't
 being filled after the RELAY_END. I really don't expect this to be
 happening.

 I tested it with an exit node and client running as separate Tor instances
 on the same machine and by the time the exit receives the RELAY_ENDs its
 cell queue for the circuit is completely full, hence the 1000 trailing
 cells the client receives after sending its last RELAY_END. There's
 nothing the exit can do to save bandwidth here except tear down the
 or_conn if it establishes all the streams on all the circuits multiplexed
 on that or_conn have received a RELAY_END. I tried this in my testing too,
 and it works well enough - but there is no pretty way of destroying an
 or_conn without flushing all the bytes on it first so I'm not sure if such
 an approach is acceptable, even if it's worthwhile.

 I'm going to work with Karsten to implement a metric for counting the
 number of bytes/cells exits flush to circuits with no active streams.

 I've attached the exit's log from the test.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/1826#comment:11>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online