- tor-relays - lists.torproject.org

No port IPv4 80 so not an called exit even when 443 open wide.
by Dr Gerard Bulger 14 Dec '20

14 Dec '20

It is now out of date that Tor servers are required to have port 80 IPv4 open, even if limited to a single Class A network in order that the relay can be labelled as an exit. Port 443 should be enough. For reasons I do not understand, if I open port IPv4 80 to a wider range I get abuse notices within a day, usually DMCA, as if some man in the middle is scanning the html! So far, wide open IPv6 port 80 attracts zero attention from those alleging abuse. http://example.com called as an IPv4 address cannot be reached by my tor exits as port 80 IPv4 too limited. Port 443 IPv4 and IPv6 on my exits are wide open to all, and we should all be on https anyway. https://example.com can be reached from my exits. Example.com does have an IPv6 address so in theory can be reached from my exits on port 80 as IPv6 port 80 is wide open I assume not used as the DNS requested for example.com gives IPv4 address. Can an Exits now be defined as an exit with port 443 open IPv4? And then tested that is an exit with https requests not http. Gerry Bulger

2 2

Re: [tor-relays] reporting Abusive exit nodes
by BRBfGWMz 13 Dec '20

13 Dec '20

It's great. Great content here too: https://wiki.archlinux.org/index.php/Tor Where do you put the configuration file when running a network from real-time distro (tails) that does not allow /etc/tor/torrc modification? On Wed, Dec 9, 2020 at 07:56 PM, lists(a)for-privacy.net <lists(a)for-privacy.net> wrote: > On 04.12.2020 05:21, BRBfGWMz wrote: > > I frequently find abusive exit nodes. Is there any way to report these > > to the > > tor consensus system, so it deprioritizes them? > You can also post here or in tor talk. bad-relays is a closed list. :-( > > > Or can I build a configuration file or list of exit nodes that I don't > > want in > > my chain? > > Of course: > > ## A list of identity fingerprints and country codes of nodes to use for the > first hop in your normal circuits. (nuts, NO address patterns!? > EntryNodes $fingerprint,{cc} > ## A list of identity fingerprints, country codes, and address patterns of > nodes to use as exit node. > ExitNodes $fingerprint,{cc},203.0.113.0/24,[2001:DB8::] > ## A list of identity fingerprints, country codes, and address patterns of > nodes to avoid when building a circuit. > ExcludeNodes $fingerprint,{cc},203.0.113.0/24,[2001:DB8::] > ## A list of identity fingerprints, country codes, and address patterns of > nodes to never use when picking an exit node. Nodes listed in ExcludeNodes > are automatically in this list. > ExcludeExitNodes $fingerprint,{cc},203.0.113.0/24,[2001:DB8::] > > ## A list of identity fingerprints, nicknames, country codes, and address > patterns of nodes that are allowed to be used as the second hop in all > client or service-side Onion Service circuits. > HSLayer2Nodes > ## A list of identity fingerprints, nicknames, country codes, and address > patterns of nodes that are allowed to be used as the third hop in all client > and service-side Onion Service circuits. > HSLayer3Nodes > > I don't have set StrictNodes 1. Please read the man page beforehand! > > Ncknames are unfortunately no longer allowed everywhere. I do not use > country codes the database has some errors. > > Soon you can use multiple lines: > > https://github.com/torproject/tor/pull/2066/commits/bb9707420e87a1efc016aea… > > \-- > p_o Ciao Marco! > > Debian GNU/Linux > > It's free software and it gives you freedom! > _______________________________________________ > tor-relays mailing list > tor-relays(a)lists.torproject.org > https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays > \-- Sent using MsgSafe.io's Free Plan Private, encrypted, online communication For everyone. https://www.msgsafe.io

3 2

My app which makes running tor relays on Windows Very Simple
by Keifer Bly 12 Dec '20

12 Dec '20

Hi all, So this is a software peice I have been developing which makes running tor relays on Windows very easy. All the user needs to do is write their torrc. The app installs tor, and configures all once the torrc is written by itself. Here is a video showing the app: Features Automatic Tor download and installation (via Choclatey) Type the torrc file, and the app configures and points tor there for you. Attempted automatic firewall configuration. Automatic creation of log file. Cons: Currently only directly support vanilla bridges. OBFS4 will have to be manually downloaded / configured. ATTEMPTED firewall auto configuration. It auto adds a rule to Windows Defender Firewall to allow the traffic, but can't configure router firewalls, etc. Let know what you think. I have successfully gotten tor up and running through this. If chocolatey changes their install scripts, this will need to be updated. I can send the source code if interested. Programming Language: C#/.Net Tor Windows App.mp4 <https://drive.google.com/file/d/1ZJ4EKXbu2Clyzc5WdNdxSWquswhdvSTJ/view?usp=…> --Keifer

1 0

reporting Abusive exit nodes
by BRBfGWMz 09 Dec '20

09 Dec '20

I frequently find abusive exit nodes. Is there any way to report these to the tor consensus system, so it deprioritizes them? Or can I build a configuration file or list of exit nodes that I don't want in my chain? \-- Sent using MsgSafe.io's Free Plan Private, encrypted, online communication For everyone. https://www.msgsafe.io

3 2

Shutdown of my exit relay
by Sarah Harvey 05 Dec '20

05 Dec '20

Hi folks, Through an error on my part, my exit relay got taken down (as12876.tor.shh.sh with fingerprint D7082DB97E7F0481CBF4B88CA5F5683399E196A3). I'm in the process of getting a replacement, but the fingerprint and IP will change as a result, and of course it will take time for it to regain consensus in the network. I believe I had also asked for it to be added to the fallback directory list since at the time it was high-traffic and stable. In this case I recommend removing it from this list until I can find a suitable replacement. Many apologies, -Sarah

1 0

Update on New Australian Law Against Domestic Hosts
by Christian Heinrich 03 Dec '20

03 Dec '20

https://www.smh.com.au/politics/federal/takeover-accounts-federal-police-to… -- Regards, Christian Heinrich http://cmlh.id.au/contact

1 0

Fwd: Thailand block
by BRBfGWMz 28 Nov '20

28 Nov '20

What provider do you use? How much does it cost? \---------- Forwarded message ---------- From: Dr Gerard Bulger <gerard(a)bulger.co.uk> Date: On Tue, Nov 24, 2020 at 08:29 AM Subject: [tor-relays] Thailand block To: BRBfGWMz <brbfgwmz(a)concealed.company> I have been running an exit (mythaicontribution) in Thailand for 3 Years (the exit has narrow port 80 range but full 443 and all other less risky ports open. It has never been that busy as an exit, unlike my other UK and USA exits of similar profile. I noticed it was off line last week. The ISP was slow to work out what was going on, to discover that the upstream ISP has blocked the IP “due to DOS attack” I think that is unlikely for such a long block, and assume the hand of MICT. I gather the do not give reasons. (https://en.wikipedia.org/wiki/Internet_censorship_in_Thailand) In some ways surprising it took them so long. Is there a way on Tor’s web pages and data base to look at historical data. I am curious to know when it off line and removed from Tor metrics. Gerry \-- Sent using MsgSafe.io's Free Plan Private, encrypted, online communication For everyone. https://www.msgsafe.io

2 1

Call for Testing - New Feature: Relay IPv6 Address Discovery
by David Goulet 28 Nov '20

28 Nov '20

Greetings everyone! We've very recently merged upstream (tor.git) full IPv6 supports which implies many many things. We are still finalizing the work but most of it is in at the moment. This is a call for help if anyone would like to test either git master[1] or nightly builds[2] (only Debian) to test for us a specific feature. The feature we would love for some of you to test is the IPv6 address discovery. In short, with this new feature, specifying an ORPort without an address will automatically bind tor to [::]:<port> and attempt to find the IPv6 address by looking at (in this order): 1. "Address" from torrc 2. "ORPort address:port" from torrc 3. Interface address. First public IPv6 is used. 4. Local hostname, DNS AAAA query. If all fails, the relay will simply never publish an IPv6 in the descriptor but it will work properly with the IPv4 (still mandatory). The other new thing is that now tor supports *two* "Address" statement which can be a hostname or IPv4 or IPv6 now. Thus this is now valid: Address 1.2.3.4 Address [4242::4242] ORPort 9001 Your Tor will bind to 0.0.0.0:9001 and [::]:9001 but will publish the 1.2.3.4 for the IPv4 address and [4242::4242] for IPv6 in the descriptor that is the address to use to reach your relay's ORPort. Now, if you happen to have this configuration which I believe might be common at the moment: ORPort 9001 ORPort [4242::4242]:9001 The second ORPort which specifies an IPv6 address will supersede the "ORPort 9001" which uses [::] and thus you will bind on 0.0.0.0:9001 and [4242::4242]:9001. You should get a notice log about this. Thus the recommended configuration to avoid that log notice would be to bind to specific addresses per family: ORPort <IPv4>:9001 ORPort <IPv6>:9001 And of course, if you want your relay to _not_ listen on IPv6: ORPort 9001 IPv4Only In your notice log, you will see which address is used to bind on the ORPort and then you will see the reachability test succeed or not on the address that tor either used from the configuration or auto discovered that is the address you are supposedly reachable from. Man page has NOT been updated yet, it will arrive once we stabilize the IPv6 feature and everything around it. Please, do report (on this thread) _anything_ even slightly annoying about this like logging or lack of logging and so on. This is a complex feature and errors can be made thus any testing you can offer is extremely appreciated. Thanks!! David [1] https://gitweb.torproject.org/tor.git/ [2] https://2019.www.torproject.org/docs/debian.html.en -- EeJVrrC/dHQXEXYB1ShOOZ4QuQ8PMnRY2XGq4BYsFq4=

8 12

Thailand block
by Dr Gerard Bulger 24 Nov '20

24 Nov '20

I have been running an exit (mythaicontribution) in Thailand for 3 Years (the exit has narrow port 80 range but full 443 and all other less risky ports open. It has never been that busy as an exit, unlike my other UK and USA exits of similar profile. I noticed it was off line last week. The ISP was slow to work out what was going on, to discover that the upstream ISP has blocked the IP "due to DOS attack" I think that is unlikely for such a long block, and assume the hand of MICT. I gather the do not give reasons. (https://en.wikipedia.org/wiki/Internet_censorship_in_Thailand) In some ways surprising it took them so long. Is there a way on Tor's web pages and data base to look at historical data. I am curious to know when it off line and removed from Tor metrics. Gerry

3 3

Fwd: Re: Bridges under DDoS
by William Kane 23 Nov '20

23 Nov '20

Forwarding to mailing list since OP has replied to my e-mail address, and not the list. ---------- Forwarded message ---------- From: William Kane <ttallink(a)googlemail.com> Date: Mon, 23 Nov 2020 16:07:39 +0000 Subject: Re: Re: [tor-relays] Bridges under DDoS To: BRBfGWMz <brbfgwmz(a)concealed.company> Dear BRBfGWMz, you can use this script I made a while ago - please read the entire message carefully or you risk locking yourself out of the system. Here you go: #!/bin/bash IPTABLES="/usr/sbin/iptables" # Flush all old rules and tables. $IPTABLES -F $IPTABLES -X $IPTABLES -t nat -F $IPTABLES -t nat -X # Establish default policies. $IPTABLES -P INPUT DROP $IPTABLES -P FORWARD DROP $IPTABLES -P OUTPUT ACCEPT # Unconditionally allow local traffic. $IPTABLES -A INPUT -i lo -j ACCEPT # Allow already established connections. $IPTABLES -A INPUT -i <NETWORK_INTERFACE> -m state --state ESTABLISHED -j ACCEPT # Allow Tor ORPort traffic. $IPTABLES -A INPUT -i <NETWORK_INTERFACE> -p tcp --dport <YOUR_ORPORT> -j ACCEPT Note that this is for a normal Tor relay, not a bridge - for a bridge, just opening the inbound obfs4proxy port is enough, though. You will have to make adjustments, namely specifying the name of your public network interface ("ip addr" or "ifconfig", either one will work on modern distributions). Make sure to also open the port sshd, your SSH daemon, bound to or you will be locked out from the system after executing the script (!). If you are on Arch Linux, you can save the script to /etc/iptables/iptables.rules (execute the script and then run "iptables-save > /etc/iptables/iptables.rules", after that the only thing left to do is to enable the according systemd service unit to automatically load the rules on boot ("systemctl enable iptables"). If on Debian, you can specify a script to run as soon as your public network interface reaches the "up" state, directly in your network configuration. Note that this will just drop unknown incoming connections (ICMP traffic as well, so you won't be able to ping your server anymore), usually the best measure, you could also configure it to reject unknown TCP packets but this will generate additional outbound traffic (which is what an attacker might want). Please also note that if you are running a IPv6 enabled relay, you will have to use ip6tables instead of iptables - you will also have to allow all incoming ICMPv6 traffic through ip6tables, with plain IPv4 you don't really need to allow ICMP packets, with IPv6, it's absolutely required. Hope I helped, never be afraid to ask questions, I am here to help. 2020-11-22 21:12 GMT, BRBfGWMz <brbfgwmz(a)concealed.company>: > Are the locations for the relays that are set up during the process of > setting > up relays for a single machine truely random (or pseudorandom), or is there > any dependence on the ip/location of my desktop? > > On Tue, Nov 17, 2020 at 08:48 AM, William Kane <ttallink(a)googlemail.com> > wrote: > >> If your server is not responding, no harm done (likely already done if > >> you have iptables set up to drop unknown (established flag not set) > >> incoming traffic.). > >> > >> If it's somehow maxing out your connection speed, then time to talk to > >> your upstream provider / hosting company - very likely they already > >> have anti dos equipment on their network, and many would just re-route > >> a customer through it until the attack stops, but it might cost you > >> some money. > >> > >> 2020-11-13 17:03 GMT, Jonas <jonasdietrich(a)ctemplar.com>: > >> > According to packetflow traces, roughly 26,000 unique IPv4 from 85 >> countries > >> > executing a SYN flood attack on TCP/9001. None of my bridges have 9001 >> open > >> > nor respond to this port. The attacks continue to this day. > >> > > >> > Danke. > >> > > >> > Jonas > >> > _______________________________________________ > >> > tor-relays mailing list > >> > tor-relays(a)lists.torproject.org > >> > https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays > >> > > >> _______________________________________________ > >> tor-relays mailing list > >> tor-relays(a)lists.torproject.org > >> https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays > >> > > \-- Sent using MsgSafe.io's Free Plan Private, encrypted, online > communication > For everyone. https://www.msgsafe.io > > >

1 0