[tor-bugs] #28598 [Core Tor/sbws]: Should torflow scaling use the consensus bandwidth when it is measured?

[Tor Bug Tracker & Wiki]

#28598: Should torflow scaling use the consensus bandwidth when it is measured?
---------------------------+-------------------------------------
 Reporter:  juga           |          Owner:  (none)
     Type:  defect         |         Status:  new
 Priority:  Medium         |      Milestone:  sbws 1.0 (MVP must)
Component:  Core Tor/sbws  |        Version:
 Severity:  Normal         |     Resolution:
 Keywords:                 |  Actual Points:
Parent ID:                 |         Points:
 Reviewer:                 |        Sponsor:
---------------------------+-------------------------------------

Comment (by teor):

 Replying to [ticket:28598 juga]:
 > sbws is not using the consensus bandwidth, but it seems that torflow
 used the consensus bandwidth instead of the descriptors bandwidth, when it
 has the measured flag
 (https://gitweb.torproject.org/pytorctl.git/tree/TorCtl.py#n376).

 Using the consensus bandwidth ignores the RelayBandwidthRate /
 MaxAdvertisedBandwidth, which is a bug in torflow.

 Here's what I suggest:
 * when the relay is unmeasured, use the minimum of all descriptor
 bandwidths (like torflow)
 * when the relay is meansured, use the minimum of the measured bandwidth,
 and the RelayBandwidthRate/MaxAdvertisedBandwidth in the descriptor

 > Ratio might need some change too
 (https://gitweb.torproject.org/pytorctl.git/tree/TorCtl.py#n1636).

 Yes, let's change the ratio to 1.0 for unmeasured relays. It's more secure
 that way.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/28598#comment:2>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online