[tor-relays] Exit probability

[Dr Gerard Bulger]

How is exit probability counted?  Is it only port 80 exit tested?  I exit many 1000s of ports, including 443, but not those of high risk of abuse emails and thus upsetting the ISP.  So port 80 along with others are blocked.  I realise no port 80 limits the use of the exit so not expecting so see a high probability of exit. But Atlas shows none.

Currently of some 3000 connections 200 are exit.

 Tor atlas show exit probability of 0.0000%

Gerry


-----Original Message-----
From: tor-relays [mailto:tor-relays-bounces at lists.torproject.org] On Behalf Of teor
Sent: 29 October 2017 12:45
To: tor-relays at lists.torproject.org
Subject: Re: [tor-relays] sum of consensus weight of 2 relays running at the same IP

On 29 Oct 2017, at 23:30, Toralf Förster <toralf.foerster at gmx.de> wrote:

>> On 10/29/2017 01:24 PM, teor wrote:
>> Possibly.
>> 
>> Are the relays CPU-limited, or bandwidth-limited?
> 
> Not at all, neither limited by a config value nor by the hardware 
> (1GBit/s, 200 MBit/s guaranteed, i7-3930, all non-Tor processes have 
> "nice" in front, ids are 1AF72E8906E6C49481A791A6F8F84F8DFEBBB2BA and 
> 6EABEBF38CE7E3DF672C4DB01383606FE3EB2215)

I'm sorry, this doesn't help me answer your question.

Usually, a relay is limited by either the available network bandwidth, or by the speed of a single CPU core on the machine.

If the first relay used all the available bandwidth, then two relays will eventually have lower consensus weight values.

If the first relay used all of a single CPU core for its main thread, then the two relays will eventually have similar consensus weight values.

T
_______________________________________________
tor-relays mailing list
tor-relays at lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays