Switching Tor relay speed

Scott Bennett bennett at cs.niu.edu
Sun Feb 22 22:27:55 UTC 2009


     On Sun, 22 Feb 2009 23:10:51 +0100 slush <slush at slush.cz> wrote:
>>     No one gets cut off when the rate is reduced.  In fact, if you've ever
>> tried reducing the rate with the BandwidthRate statement in torrc, you may
>> have noticed that the actual data rate takes a while to slow down to the
>> newly
>> imposed limit.
>>
>
>I tried that today on my laptop relay with vidalia. Speed change is applied
>in few seconds.

     Are you sure that wasn't just coincidence?  Rates zoom up and down all
the time.  It has been at least a year and a half since I last needed to do
that, but at that time it typically took several minutes to get the rate down
to somewhere near the new limit.  Even then, it was very inconsistent for a
while, perhaps until a large fraction of active clients had picked up the new
version of the descriptor.  I noticed further that BandwidthBurst was
frequently exceeded by a sizable percentage, which, as I read the man page,
should not happen.
>
>But again, when I slow down speed from 50 to 20 kB/s, it isn't so
>problematic like 3000 and 100 kB/s, as we spoke before.

     Well, it appears that the relay itself doesn't do a lot of enforcement
of limits.  Instead, it is the population of clients that simply choose not
to build circuits through your relay when it advertises low rates.  If you
look at the torstatus page at just about any time, you will see that there
are far more relays at the slow end of the rate spectrum than toward the
fast end, so the chances of a particular slow relay being selected for a
circuit's route are smaller than for a fast route, just as it should be.
Fewer circuits through your relay mean fewer chances that a client or relay
somewhere will need to connect to your relay in order to carry a circuit.
Having only, say, 200 - 300 connections instead of 2,000 - 3,000 connections
is part of the effect you're seeing.
>
>Marek
>
>--00163630f629af8d520463892804
>Content-Type: text/html; charset=ISO-8859-1
>Content-Transfer-Encoding: quoted-printable
>
><div class=3D"gmail_quote"><blockquote class=3D"gmail_quote" style=3D"borde=
>r-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-le=
>ft: 1ex;">&nbsp; &nbsp; No one gets cut off when the rate is reduced. &nbsp=
>;In fact, if you&#39;ve ever<br>
>
>tried reducing the rate with the BandwidthRate statement in torrc, you may<=
>br>
>have noticed that the actual data rate takes a while to slow down to the ne=
>wly<br>
>imposed limit.<br></blockquote></div><br>I tried that today on my laptop re=
>lay with vidalia. Speed change is applied in few seconds.<br><br>But again,=
> when I slow down speed from 50 to 20 kB/s, it isn&#39;t so problematic lik=
>e 3000 and 100 kB/s, as we spoke before.<br>
><br>Marek<br>
>
>--00163630f629af8d520463892804--
>
     You really don't need to send an HTML copy of your message to the list.


                                  Scott Bennett, Comm. ASMELG, CFIAG
**********************************************************************
* Internet:       bennett at cs.niu.edu                              *
*--------------------------------------------------------------------*
* "A well regulated and disciplined militia, is at all times a good  *
* objection to the introduction of that bane of all free governments *
* -- a standing army."                                               *
*    -- Gov. John Hancock, New York Journal, 28 January 1790         *
**********************************************************************



More information about the tor-talk mailing list