[tor-bugs] #33832 [Core Tor/sbws]: For relays that change ip, only the measurements with the last ip are kept

Tor Bug Tracker & Wiki blackhole at torproject.org
Sat Apr 11 04:42:57 UTC 2020


#33832: For relays that change ip, only the measurements with the last ip are kept
-------------------------------------------------+-------------------------
 Reporter:  juga                                 |          Owner:  gk
     Type:  defect                               |         Status:
                                                 |  needs_information
 Priority:  Medium                               |      Milestone:  sbws:
                                                 |  1.1.x-final
Component:  Core Tor/sbws                        |        Version:  sbws:
                                                 |  1.1.0
 Severity:  Normal                               |     Resolution:
 Keywords:  sbws-roadmap, intro,                 |  Actual Points:
  GeorgKoppen202004                              |
Parent ID:  #33121                               |         Points:
 Reviewer:                                       |        Sponsor:
-------------------------------------------------+-------------------------

Comment (by gk):

 Replying to [comment:5 juga]:
 > Replying to [comment:4 gk]:
 > > Replying to [comment:2 juga]:
 > > > To solve this, the function `trim_results_ip_changed` in
 `resultdump.py` should be changed
 > >
 > > What is the intended behavior? Not ignoring previous measurements
 anymore? Do we still need the `trim_results_ip_changed` function in that
 case (you seem to indicate that)?
 >
 > Hmm, i thought that we could ignore the other ips but still keep the
 results. After your questions, i realize that it'd not make sense to do
 that.
 > So, we can probably close this ticket as invalid. Maybe let fix other
 tickets first and then rethink about this one?

 We can do that. But why do we keep only the measurements just with the
 last IP address? I mean it's still the *same* relay, just has a new IP
 address (you could probably argue that the IP address is just an attribute
 like an option in the `torrc` file and we do not exclude measurements when
 any of those changes). Is the assumption that bandwidth improves with a
 new IP address? Or is Torflow doing that and sbws is trying to mimic it?
 Or...?

 I looked at the `git` history of this code to figure out whether there is
 some rationale given in a commit message or ticket it points to or some
 older comment but did not find any hint.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/33832#comment:6>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online


More information about the tor-bugs mailing list