Re: [tor-talk] Default clients to be non-exit relay LibTech x

[Moritz Bartl <moritz@xxxxxxxxxxxxxx>]

On 28.08.2013 04:30, Percy Alpha wrote:
>> Theoretically, the directory
>> authority could serve up a set of a thousand random nodes, and then the
>> client software could select a random set to use initially.
> This will make blocking tor nearly impossible and eliminate the need for
> Bridge node since authority will have a hard time keeping up with all
> non-exit node dynamical.

In over 10 years of development and thinking going into Tor, do you
think you're the first to come up with this idea? Nobody is stopping
anyone from coming with a proper design, but it's not as easy as you
seem to believe.

Every client has to download the full list of relays ("consensus")
periodically. In areas with little connectivity, this already puts a
high burden on clients.

It is very hard for Tor clients to determine which relays to use. If you
have many low bandwidth relays, you need a good way to determine if a
relay is currently "available", or already in use.

A related ticket is https://trac.torproject.org/projects/tor/ticket/1854

"Mike's performance work has shown that the smaller relays -- for
example, the ones that set bandwidthrate and bandwidthburst to 20k --
are never good news to have in your circuit. Damon McCoy's hotpets 2010
paper showed more details of how you could improve performance by
dumping the bottom X% of the relays."

A good document to read is the Tor Path Specification:
https://gitweb.torproject.org/torspec.git/blob/HEAD:/path-spec.txt

I'm not saying that it's impossible to make use of low bandwidth relays,
but it would need a careful design and new path selection mechanism.

-- 
Moritz Bartl
https://www.torservers.net/
-- 
tor-talk mailing list - tor-talk@xxxxxxxxxxxxxxxxxxxx
To unsusbscribe or change other settings go to
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-talk