[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: [tor-bugs] #31518 [Core Tor/Tor]: HAProxy implementation in TCPProxy option.



#31518: HAProxy implementation in TCPProxy option.
--------------------------+----------------------------------
 Reporter:  haxxpop       |          Owner:  haxxpop
     Type:  enhancement   |         Status:  needs_revision
 Priority:  Medium        |      Milestone:  Tor: unspecified
Component:  Core Tor/Tor  |        Version:
 Severity:  Normal        |     Resolution:
 Keywords:  proxy tcp     |  Actual Points:
Parent ID:                |         Points:
 Reviewer:  nickm         |        Sponsor:
--------------------------+----------------------------------

Comment (by haxxpop):

 Replying to [comment:5 nickm]:
 > First question: on your note above, what do you mean by "the proxy has
 to support DNS resolving in HTTP CONNECT"?  We *want* all the DNS
 resolution to happen in Tor. Tor should really only be passing IP
 addresses; does it pass hostnames to HTTP CONNECT proxies?

 Yes Tor will be passing only IP addresses.

 I meant that if I want to deploy my own proxy server and if it's HTTP
 CONNECT, it's supposed to support DNS resolving because it will be
 probably used by some other proxy client (which is not Tor). So it's more
 reasonable for me to deploy a haproxy instead of HTTP CONNECT proxy.



 > Second question: is it actually a problem in practice that SOCKS5 takes
 two round trips?
 I'm not sure. I don't have much experience on this :D. But in theory,
 according to the speed of light, if the server is in the US and the client
 is in Thailand, it will take at least 80ms more if it's two round trips.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/31518#comment:9>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs