[tor-talk] Fwd: Exit Enclaves (Roger Dingledine)

Zach Lym indolering at gmail.com
Wed Apr 27 04:25:19 UTC 2011


> > I am doing page load tests on an exit enclave (duckduckgo.com) over a
> period
> > of a few weeks.  I am worried that I will exit at other nodes and reach
> the
> > public duckduckgo site.
>
> % host duckduckgo.com
> duckduckgo.com has address 50.17.207.245
>
> % grep 50.17.207.245 cached-*
> %
>

Odd, I will CC the CEO.


> I think there is no Tor relay running at that IP address currently,
> so this exit enclave you're hoping for cannot happen. An exit enclave
> is a Tor exit relay that runs on the same IP address as a destination
> (e.g. a website).
>
> > The documentation is somewhat ambiguous on this point: If a service
> provides
> > an enclave node, the standard Tor out-proxy's are *not* used, correct?
>
> Perhaps you mean a Tor hidden service, not an exit enclave?
>

No, I mean a properly configured exit enclave which (as I understand it) is
reachable when using Tor without using out-proxies. Out-proxies are the real
choke-point of the Tor network...


> In which case, yes, if you visit a .onion address, that won't involve
> Tor exit relays. You instead do a rendezvous inside the Tor network:
> https://www.torproject.org/docs/hidden-services
>

> > Thank you,
> > > P.S. Would using paid proxy's which exit in other continents improve
> > external validity?
>

> What does 'external validity' mean?
>
Similarity to the real world and/or applicability outside of the testing
environment.  So, would someone in China or Europe get differen circuits, is
using the same IP and Tor node over and over going to create caches, etc.


> --Roger
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.torproject.org/pipermail/tor-talk/attachments/20110426/16bf6401/attachment.htm>


More information about the tor-talk mailing list