[tor-bugs] #2286 [Tor Directory Authority]: We still use self-published relay bandwidth sometimes

[Tor Bug Tracker & Wiki]

#2286: We still use self-published relay bandwidth sometimes
-------------------------------------+--------------------------------------
 Reporter:  arma                     |          Owner:                    
     Type:  defect                   |         Status:  needs_review      
 Priority:  major                    |      Milestone:  Tor: 0.2.3.x-final
Component:  Tor Directory Authority  |        Version:                    
 Keywords:  arma-cares               |         Parent:                    
   Points:                           |   Actualpoints:                    
-------------------------------------+--------------------------------------

Comment(by arma):

 Looks like it should work. Three thoughts:

 A) I notice that this plan doesn't affect flags at all. So we could have a
 relay with the Fast flag but a weight of 20 (which would ordinarily not
 warrant the Fast flag). That's a separate topic (whether authorities
 should vote flags based on the descriptor bandwidth, the vote bandwidth,
 or now the consensus bandwidth), but I'd be more comfortable if we raise
 our default cap to e.g. 100.

 B) It would seem this change requires a corresponding change in the
 torflow bwauth code, to not admit opinions about any relays until it has
 'enough' opinions. Or maybe torflow is the wrong place to put that logic,
 and it should go in the Tor dir auth logic, either in
 dirserv_read_measured_bandwidths() or probably better in
 routerstatus_format_entry() when we're deciding
 {{{
     if (format == NS_V3_VOTE && rs->has_measured_bw) {
 }}}

 C)
 {{{
 -    "w" SP "Bandwidth=" INT [SP "Measured=" INT] NL
 +    "w" SP "Bandwidth=" INT [SP "Capped=1"] [SP "Measured=" INT] NL
 }}}
 Is the order mandatory here? Sure looks like it is. I guess in our current
 implementation it doesn't matter, since we never say Capped and Measured
 together.

 Smaller thoughts:

 A) run make check-spaces on it

 B) the spec change says "maxunmappedbw" which is probably not the right
 word.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/2286#comment:33>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online