August 2016 - tor-relays - lists.torproject.org

Re: [tor-relays] Cheapest HW to get 20Mbit?
by Joel Cretan 29 Aug '16

29 Aug '16

Having run a relay on an older RPi with standard Raspbian, I would caution you to look carefully at the packages you're using, if you choose that hardware. Of course the Tor package itself is woefully out of date, so you have to build from source. But it's worse than that. I noticed that running regular apt-get update && apt-get upgrade was not enough to keep openssl up to date. Over a year after Heartbleed had been fixed, I noticed that my "up-to-date" version was still vulnerable, not to mention all the other bugs discovered there in the last few years. I thought maybe I could replace openssl with one of the forks, but was unable to find any pre-built packages or even signed source distributions with signing keys distributed over TLS. It was a pretty bad state of affairs, so I shut that relay down entirely. Maybe it has improved since then, but be careful. On Aug 28, 2016 4:37 PM, "Matt Traudt" <sirmatt(a)ksu.edu> wrote: On 08/28/2016 04:26 PM, Petrusko wrote: > >> Up to two per IP. > Hu? it's sad for people having several CPU... :s > It does help a little to prevent attackers from spinning up a lot of relays. With this limit, they must have n/2 IPs at their disposal. For example, this paper[1] shows an attack for harvesting onion services. It would have been much easier without the 2-per-IP limit. Matt [1]: http://ieee-security.org/TC/SP2013/papers/4977a080.pdf _______________________________________________ tor-relays mailing list tor-relays(a)lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

2 1

Closing a relay, to move/upgrade, identity question ?
by Petrusko 29 Aug '16

29 Aug '16

Hey, Planning to close, then re-build a relay. (1st time I'm testing it on the whole system!) Some useful informations here about : https://www.torproject.org/docs/faq.html.en#UpgradeOrMove I think it's useful to stay on the same "identity"... But if the new torrc will use others TCP ports, will it be a mess ? No problem for me if this (new) relay will not have the same identity has before. But it will be nice to retore Atlas old graphs ;) If I'm paranoiac, and if this current relay has been corrupted, I think it's better to start a clean identity without the old keys ? Thx for your lights ;) -- Petrusko PubKey EBE23AE5 C0BF 2184 4A77 4A18 90E9 F72C B3CA E665 EBE2 3AE5

4 4

block input hammering from the same ip source address
by Toralf Förster 29 Aug '16

29 Aug '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 These are iptables rules (ipv4) for my exit relay: IPT="/sbin/iptables" # Tor # $IPT -A INPUT -p tcp --destination-port 80 --match conntrack --ctstate NEW --match connlimit --connlimit-above 2 --connlimit-mask 32 -j DROP $IPT -A INPUT -p tcp --destination-port 443 --match conntrack --ctstate NEW --match connlimit --connlimit-above 2 --connlimit-mask 32 -j DROP # $IPT -A INPUT -p tcp --destination-port 80 -j ACCEPT $IPT -A INPUT -p tcp --destination-port 443 -j ACCEPT For the first 2 I do wonder if there's something I should consider too ? - -- Toralf PGP: C4EACDDE 0076E94E, OTR: 420E74C8 30246EE7 -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iF4EAREIAAYFAlfEh1wACgkQxOrN3gB26U4C2QD/Uvptq50hLXDbgLBwfKOWX60D 0ezIAtlBiXdjI86IeZkA/0ZW847biWDw3e51xLZpxGskA1aoD2mhz+xIbXvFMS/h =ZFV5 -----END PGP SIGNATURE-----

1 0

tor 0.2.8.x: in 26 days to the top
by nusenu 29 Aug '16

29 Aug '16

~26 days after the first stable tor 0.2.8.x release, it has become the most prevalent tor version on the network by consensus weight [1]. There are still more 0.2.7.x relays than 0.2.8.x relays on the network [2], but that will change soon. [1] https://github.com/ornetstats/stats/blob/9f63665a7a4bc996f224006e76b1c7cf4e… [2] https://metrics.torproject.org/versions.html

1 0

Cheapest HW to get 20Mbit?
by Daniel Armyr 28 Aug '16

28 Aug '16

Hi Since a few months back I have two RPi running Tor relays and delivering about 1Mbit each of peak throughput. Now my ISP has increased my max upload speed and I am looking to contribute more, somewhere in the 20Mbit range or possibly more. But what hardware do I need? Has anyone written down a list of minimum requirements to get a certain throughout? I have tried to google, but failed. Sincerely Daniel Armyr.

3 4

Re: [tor-relays] Cheapest HW to get 20Mbit?
by Pi3 28 Aug '16

28 Aug '16

Hi Daniel, Judging from my pi3 mid relay where 5mbit traffic is about 20% of 1 core (tor uses 1 core mostly) 20-25mbit should be fine for pi3. In the peak hours it will be probably close to maximum what pi3 is capable on 1 core but when cost is a factor I would not think of anything else 1gb ram is more than enough, my tor uses about 150mb If you think of exit, cpu load may be higher

1 0

Re: [tor-relays] interesting network sockets pattern
by Toralf Förster 28 Aug '16

28 Aug '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 08/25/2016 07:02 PM, Toralf Förster wrote: > This is a fresh new Tor exit, setup 4 days ago, https://atlas.torproject.org/#details/BE2FA9FCB6242567B93ED99FEC5543FC517C9… , where I do wonder how to interpret the attached screen shot. The full SVG graphic can be derived from http://www.zwiebeltoralf.de/torserver/graph-20160825.svg (scroll down to the end) - -- Toralf PGP: C4EACDDE 0076E94E, OTR: 420E74C8 30246EE7 -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iF0EAREIAAYFAlfAOBwACgkQxOrN3gB26U7NEAD/U8eOOYq+sArqaGdrJ5+RzTUY G8duzPfhX6bbiI2u55cA93IBBN/7Hog4WhftxhM681ntxKk44KkYOryqyIaBZp4= =WWe1 -----END PGP SIGNATURE-----

4 7

Re: [tor-relays] interesting network sockets pattern
by Jivan Amara 27 Aug '16

27 Aug '16

> On Friday 26 August 2016 14:37:48 Toralf Förster wrote: >> On 08/25/2016 07:02 PM, Toralf Förster wrote: >>> This is a fresh new Tor exit, setup 4 days ago, >>> https://atlas.torproject.org/#details/BE2FA9FCB6242567B93ED99FEC5543FC517 >>> C9276 , where I do wonder how to interpret the attached screen shot. >> The full SVG graphic can be derived from >> http://www.zwiebeltoralf.de/torserver/graph-20160825.svg (scroll down to >> the end) > Hi Toralf, > > I see the same patterns on my relays from time to time. Please see my 2012 > posting https://lists.torproject.org/pipermail/tor-relays/2012-December/001781.html on [tor-relay] on that subject. I still have no idea what > causes the pattern. > > Regards, > > torland > I suspect this is due to a circuit built with a node that has lower max bandwidth settings than yours. The regular bandwidth cap holds traffic back for a time, then the burst bandwidth cap lets some through at a higher speed for a short period. -J

2 1

GeoIP
by Fred Rauch 27 Aug '16

27 Aug '16

Hi, I just started up another relay, and atlas' GeoIP data on it is incorrect (says it is in Netherlands when it is in fact in Lithuania). Where does Atlas get its GeoIP data from? Can this be corrected? It's not a huge issue but it would be nice for the correct country to be displayed. albino

3 2

Some information about Tor relays
by Liu, Zhuotao 26 Aug '16

26 Aug '16

Hi Folks, This is Sky from University of Illinois. Currently we are working on research project related with Tor. To help us to better design and evaluation our proposal, we need some information about the Tor relays that is currently unavailable from the Atlas. Thus, if someone who operates Tor relays could provide us such information, your help will be greatly appreciated. :) We hope to have an estimate about computation capacity of Tor relays. For instance, how many circuits a relay can maintain when its CPU is driven to about 100%? On average, how many circuits are maintained by a busy guard and what the CPU utilization is. These kinds of information would be really helpful. Further, in Sep 2013, Tor saw a spike in its clients. https://blog.torproject.org/blog/how-to-handle-millions-new-tor-clients The suspicion is that a botnet used Tor as its C&C channel. Is there any data about the relay utilization during that time? For instance, how many more circuits a relay had to maintain during that time? Thanks a lot for your help. :)

2 2