- tor-relays - lists.torproject.org

Question
by Андрей Гвоздев 23 Mar '21

23 Mar '21

What does it mean "Tor's file descriptor usage is at 90%. If you run out Tor will be unable to continue functioning."?

2 1

Questing regarding Team Cymru Tor Relays and Bridges
by Lisa Winter 22 Mar '21

22 Mar '21

Hello list, I just visited BridgeDB and got a bridge from "Team Cymru", according to the whois of the IP - should the Tor Project really allow a company trying to "track and take down threat actors and criminals around the globe" host a substantial portion of the network, according to their info page: https://team-cymru.com/company/ The following was also very concerning: "Team Cymru is comprised of former…" - ... - Law enforcement - ISP backbone engineers - ... I suspect that they log connections to their relays and bridges, and maybe even more. Such companies profit from gathering and selling information's, exploits and the likes. bad-relays rejected this message, which is concerning.. I hope someone here could redirect it to the right people, or do the Tor Project higher-ups know the people behind Team Cymru and vouch for them? Bye, Lisa Winter

3 3

Sunsetting C3WLittleGem and C3WPhantomStreak
by MacLemon 18 Mar '21

18 Mar '21

Hi! FYI: We sadly have to sunset two of our relays due to persistent network/upling problems with the vSphere cluster of our hosting provider. It wasn't possible to solve these issues in about a year. To reduce our volunteers' time investment for no gain, we'll turn of C3WLittleGem and C3WPhantomStreak in the next few days. If you want to have a last look at them, here's a link to our family. https://metrics.torproject.org/rs.html#search/family:B1F09810C4F7B9D30F483D… We'll continue running C3WEyeOfSauron and hope being able to add more relays again in the future. Best regards MacLemon (on behalf of Chaos Computer Club Wien (C3W))

2 1

Re: [tor-relays] Did 'Sandbox 1' break Tor for anyone else on 0.4.5.6?
by William Kane 17 Mar '21

17 Mar '21

Hi Peter, >Would be great if you could get details about the failing call. I already thought of gathering said details by tracing the process, but did not want to risk my uptime statistics, which would inevitably happen if I had to restart the server and service over and over (I disabled tracing globally through the Yama LSM as a security measure, i.e. kernel.yama.ptrace_scope == 3) - recently I lost the guard flag multiple times, caused by some sort of attack that I already reported on this list (tor-relays) - someone kept creating a fuckton of circuits through my relay (averaging 90k per minute), thus causing tor to run out of memory / get oom-killed by the kernel before it could even step in and close the circuits - if it was even trying to, it would make sense for the DoS mitigation code to be active only for the first link in the circuit aka the guard, and my node simply being a middle-only relay, it got completely stomped by said attack. After somewhat mitigating this attack by tweaking MaxMemInQueues, creating a bigger swap file and tuning vm.swappiness, I regained the guard flag, but then the hypervisor my KVM box is running under experienced some issues and had to be rebooted - once again, I received no notice of that until the relay was already offline for a few days, causing me to lose the guard flag again. Seems like luck is just not on my side these days, or well, it's been weeks now. > You should simply see a Permission Denied if the capability is the problem. Here's a copy from stdout, only happening if Sandbox is set to 1.: Mar 15 20:15:20.000 [notice] Configured to measure statistics. Look for the *-stats files that will first be written to the data directory in 24 hours from now. Mar 15 20:15:21.000 [warn] fstat() on directory /var/lib/tor_debug failed. Mar 15 20:15:21.000 [err] Can't create/check datadirectory /var/lib/tor_debug Mar 15 20:15:21.000 [err] Error initializing keys; exiting Running it as a privileged user does not change thing, so no permissions issue: Mar 15 20:17:24.000 [notice] Configured to measure statistics. Look for the *-stats files that will first be written to the data directory in 24 hours from now. Mar 15 20:17:24.000 [warn] You are running Tor as root. You don't need to, and you probably shouldn't. Mar 15 20:17:25.000 [warn] fstat() on directory /var/lib/tor_debug failed. Mar 15 20:17:25.000 [err] Can't create/check datadirectory /var/lib/tor_debug Mar 15 20:17:25.000 [err] Error initializing keys; exiting I've traced down the origin of the fstat() call to this piece of code: https://github.com/torproject/tor/blob/master/src/lib/fs/dir.c#L158 However, looking at the code that establishes and populates seccomp rules, it seems like fstat and it's 64 bit counterpart are not subject to (parameter) filtering, i.e. seccomp_rule_add_0 is invoked with the parameter SCMP_ACT_ALLOW, reading the manpage for seccomp_rule_add(3) reveals: "The seccomp filter will have no effect on the thread calling the syscall if it matches the filter rule." References: https://github.com/torproject/tor/blob/master/src/lib/sandbox/sandbox.c#L148 https://github.com/torproject/tor/blob/master/src/lib/sandbox/sandbox.c#L15… https://man7.org/linux/man-pages/man3/seccomp_rule_add.3.html So, even though technically, seccomp should allow these syscalls to be invoked, no matter which parameters are passed, somehow enabling the whole sandbox subsystem still breaks fstat. We are getting into tor-dev@ territory here, so maybe someone who is more experienced with the code base could figure out why this is happening, I'm pretty good at C/C++ on Linux / Windows and all the shenanigans that come with it, but lack time to debug this further. I've added tor-dev(a)lists.torproject.org as a CC just in case. - William 2021-03-15 17:10 GMT, Peter Gerber <tor(a)arbitrary.ch>: > Hi William > > William Kane: >> Hi everyone, >> >> Ever since I upgraded to tor version 0.4.5.6, enabling tor's built-in >> seccomp sandbox completely breaks tor, i.e. it gets killed by the >> kernel on start for a seccomp violation (fstat(..)) - sandboxing >> worked fine on 0.4.4.6, my system configuration did not change between >> the updates. > > Tor itself usually fails with a Permission Denied error when a syscall > fails due to seccomp. So, this is rather odd. > >> I figured this was happening because I do not grant the >> CAP_DAC_READ_SEARCH capability, but I'm not so sure anymore if that's >> the reason. > > You should simply see a Permission Denied if the capability is the problem. > > Would be great if you could get details about the failing call. If > seccomp is involved, you should be able to get details like this: > > • install package auditd > • make sure auditd is running > • crash Tor > • find the syscall with `ausearch -ts recent -i` > > Peter >

2 2

Did 'Sandbox 1' break Tor for anyone else on 0.4.5.6?
by William Kane 16 Mar '21

16 Mar '21

Hi everyone, Ever since I upgraded to tor version 0.4.5.6, enabling tor's built-in seccomp sandbox completely breaks tor, i.e. it gets killed by the kernel on start for a seccomp violation (fstat(..)) - sandboxing worked fine on 0.4.4.6, my system configuration did not change between the updates. Not specifying 'Sandbox 1' in the torrc allows tor to fully start as usual, I had to disable it post-update in order to be able to continue running my relay. /etc/tor/torrc: ORPort 37.157.195.83:38619 ORPort [2a02:2b88:2:1::3239:0]:38619 Nickname michaelscott ContactInfo ttallink(a)googlemail.com ControlPort 9051 SocksPort 0 CookieAuthentication 1 ExitPolicy reject *:* DataDirectory /var/lib/tor # Sandbox 1 ShutdownWaitLength 120 MaxMemInQueues 672MB /usr/lib/systemd/system/tor.service: [Unit] Description=Anonymizing Overlay Network After=network.target [Service] User=tor Type=simple ExecStart=/usr/bin/tor -f /etc/tor/torrc ExecReload=/usr/bin/kill -HUP $MAINPID KillSignal=SIGINT LimitNOFILE=32768 PrivateDevices=yes [Install] WantedBy=multi-user.target /etc/systemd/system/tor.service.d/override.conf: [Service] PrivateTmp=yes ProtectHome=yes ProtectSystem=strict ProtectKernelLogs=yes ProtectKernelModules=yes ProtectKernelTunables=yes ProtectControlGroups=yes NoNewPrivileges=yes RestrictSUIDSGID=yes RestrictAddressFamilies=AF_INET AF_INET6 ReadWritePaths=/var/lib/tor CapabilityBoundingSet=CAP_NET_BIND_SERVICE TimeoutStopSec=2min 15s I'm running ArchLinux with the 5.11.4 kernel, Tor is still sandboxed by systemd so this is not a huge issue but I would like to be able to continue using the built-in sandbox as well as systemd's sandboxing options in order to maximize process isolation. Did anyone else run into this? I would have posted a bug report but due to various reason I am not able to do so right now. I figured this was happening because I do not grant the CAP_DAC_READ_SEARCH capability, but I'm not so sure anymore if that's the reason. Any help is greatly appreciated. - William

2 1

Re: [tor-relays] I was banned from PayPal
by potlatch 13 Mar '21

13 Mar '21

Hello kantorkel and Tor ops, To put a little meat on the bone of this situation, I did not receive an email from PayPal--I was informed that I was banned when I tried to make a payment for VPS hosting. During the week before this ban, I had sent funds to both Flokinet for an exit relay in Finland and to ITLDC.com for VPS hosting in Poland--both exit relays. I used the PayPal support email to ask for the reason I was banned and to ask for re-instatement. I have heard nothing back. I had surmised that someone was monitoring payments to off-shore hosting companies and maybe they assumed I was supporting a news agency--I don't know really. If that is the case, why haven't the hosting sites been banned as well? The message in my PayPal account referred to a violation of the user agreement. Here is the full text of the message: -We're unable to continue offering our services -Thank you for using PayPal as your payment partner. -Unfortunately, we are unable to continue offering our services -to you at this time due to the nature of your business and/or -activity in your account and the risk it poses to PayPal. This -decision can't be overturned. -If you owe refunds to any of your buyers, you can use the money -in your PayPal account to refund them. -Money in your PayPal account will be held for 180 days. After -180 days, we'll email you information on how to withdraw your -money. The only transactions on this account over the past 3-months were VPS lease payments off-shore. I looked up the user agreement to determine where I had violated the rules and found only a single reference to operating servers: -Service Requiring Pre-Approval: -Providing file sharing services or access to newsgroups; or -selling alcoholic beverages, non-cigarette tobacco products, --e-cigarettes or prescription drugs/devices. I have a zoom meeting with two attorneys from EFF.org on March 18. Stay tuned....and maybe ban PayPal. --potlatch Sent with ProtonMail Secure Email. ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐ On Thursday, March 11, 2021 12:06 AM, Matthias Marx <kantorkel(a)hamburg.freifunk.net> wrote: > Hi Potlatch, > > would you mind sharing PayPal's email / full message with me? > > Regards from Hamburg > kantorkel > > Am 3/11/21 um 12:42 AM schrieb potlatch: > > > Today I received a message from PayPal that paying for Tor relay server leases was a direct violation of my usage agreement. I have been paying off-shore VPS hosts for my Tor server leases with PayPal for at least ten years. Very interesting that they act now. > > They specifically state the ban was for: > > Providing file sharing services or access to newsgroups; or selling alcoholic beverages, non-cigarette tobacco products, e-cigarettes or prescription drugs/devices. > > They further state that the ban is permanent and not reversible. > > I liked using PayPal because it was a safe way to pay and didn't expose my credit card information. Before PayPal I used a prepaid card which I will probably go back too. > > --Potlatch > > Sent with ProtonMail https://protonmail.com Secure Email. > > > > tor-relays mailing list > > tor-relays(a)lists.torproject.org > > https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

3 2

Fallback EOL
by Anders Burmeister 13 Mar '21

13 Mar '21

We lost Fallback relay 775B0FAFDE71AADC23FFC8782B7BEB1D5A92733E Had it’s EOL due to a major incident in data center. We are all sorry Regards /anders

1 0

I was banned from PayPal
by potlatch 12 Mar '21

12 Mar '21

Today I received a message from PayPal that paying for Tor relay server leases was a direct violation of my usage agreement. I have been paying off-shore VPS hosts for my Tor server leases with PayPal for at least ten years. Very interesting that they act now. They specifically state the ban was for: Providing file sharing services or access to newsgroups; or selling alcoholic beverages, non-cigarette tobacco products, e-cigarettes or prescription drugs/devices. They further state that the ban is permanent and not reversible. I liked using PayPal because it was a safe way to pay and didn't expose my credit card information. Before PayPal I used a prepaid card which I will probably go back too. --Potlatch Sent with [ProtonMail](https://protonmail.com) Secure Email.

10 9

Thoughts and insight before bridge moving
by a tor op 09 Mar '21

09 Mar '21

Hello I was thinking of perhaps asking and checking if the list has some comments on thoughts I have. I'm attaching a picture, hope that's not a problem, guess it's better than providing a URL to a file upload place (or maybe that's ok too?). I've been running a TOR bridge node for some years and need to rebuild/move it. For some time I wasn't sure how useful it was, whether or not it contributed in a useful way and I don't think it still has all the flags it could/should have? So, for you around here with an even deeper experience of the network, looking at the stats, are there any particular insights that can be drawn from there? Is it a useful bridge, stats that could or should look differently, anything else? BTW, there's a huge spike in connections on 9 Aug -20 (and there's been similar occasions earlier too) does that coincide with known attacks on the network or what? I guess, should I want to continue providing service, that it's useful or possibly preferred to move over the keys/config (?) to indicate same admin (and for users if there's no noticeable interruption the move won't even be noticed I suppose). TIA, a tor op

4 4

Bridge operator iat_mode setting
by torjoy 04 Mar '21

04 Mar '21

Hi All, I work with time and frequency references and run some tor bridges. What is the objective of "iat_mode" setting? Is an good timming reference important for this setting? For now, i'm adminstrating 3 briges, one with iat_mode=0, iat_mode=1 and iat_mode=2. Could you explain or forward me to some reading about it? Best regards, Luiz Sent with [ProtonMail](https://protonmail.com) Secure Email.

9 14