February 2026 - tor-relays - lists.torproject.org

Post-Quantum Cryptography in Tor's TLS Layer: Help needed!
by Alexander Hansen Færøy 27 Feb '26

27 Feb '26

Hello Tor Relay Operators, With the arrival of OpenSSL 3.5.0 in 2025, we finally had an important missing piece of infrastructure to start enabling the Tor network to support Post-Quantum Cryptography in the outermost cryptographic layer of the Tor protocol. Since Tor version 0.4.8.17, we have supported the x25519/ML-KEM-768 hybrid handshake in our TLS layer, and we have gradually seen more relays supporting it in the production network. Tor Browser has been built against OpenSSL >= 3.5.0 since (at least) version 13.5.22, released in September 2025, and therefore supports Post-Quantum TLS handshakes today, but it requires that its guard node supports it. We need to give a big shout-out to the OpenSSL Team here to get this out and available to the masses! <3 Since Q3 2025, I've been running some semi-regular scans of the Tor production network to check for TLS cipher suite availability. This work started as part of the first Tor Community Gathering back in October, 2025[1]. The results are as follows: In January, around 28% of the Tor relays supported the Post-Quantum TLS handshake. The scan I did yesterday showed an increase of around 4 percentage points to 32.32%. Out of 9476 scanned relays, 9185 were reachable, 291 were unreachable. Of the 9185 reachable relays, I saw the following distribution of cipher suites: `TLS13_AES_256_GCM_SHA384`: 9023 relays. `TLS13_CHACHA20_POLY1305_SHA256`: 158 relays. `TLS13_AES_128_GCM_SHA256`: 4 relays. For the key exchange groups, I saw the following distribution: `secp256r1`: 6212 relays. `X25519MLKEM768`: 2969 relays. `X25519`: 4 relays. Of the Directory Authorities, 5 of them supported the `X25519MLKEM768` key exchange group during the TLS handshakes. 2969 / 9185 * 100 = 32.32% of relays support the `X25519MLKEM768` cipher suite. This number is a good start, but we should ideally bump it up! You can see the individual results for your relay(s) in the big table available at https://ahf.me/tor-tls-pqc/2026-02-26/ -- entries without a cipher suite or key exchange group were unreachable at the time I ran the scan. These missing entries can be due to a network problem anywhere on the path between my host and yours, so don't get worried if your relay is missing :-) The scanner I used is free software and is available as part of the Network Health Team's Margot tool[2], used for various analysis purposes. Please note that the goal here isn't to enumerate all available cipher suites for each relay, but instead I'm interested in the "strongest" possible cipher suite my custom client can negotiate with each relay. My ask for you as the relay operator community here is as follows: Next time you folks are doing maintenance on your relay(s), please have a look at whether you can upgrade either your packages or underlying software distribution to a version such that you either get a new enough version of OpenSSL (>= 3.5.0), or if you are a LibreSSL user, check whether the LibreSSL project supports the ML-KEM hybrid handshake, and upgrade to that version. It looks like the LibreSSL project is currently working towards supporting the ML-KEM handshake in the next upcoming release[3]. For Debian users, to get a new enough OpenSSL version that supports the PQC TLS handshake, you need to upgrade to (at least) Debian Trixie. For everybody else, it's best to check your respective software distribution's package management system to find the version you need to upgrade to. For Arti users, a licensing issue with the currently available crypto providers in Arti prevented enabling the ML-KEM handshake. Nick, however, has recently been making progress on this matter and has a patch up that is now pending upstream review[4]! :-) If you are excited about exploring different aspects of the Tor ecosystem, talking with other Tor enthusiasts, and you have nothing planned for your weekend in two weeks, I highly encourage you to look into the next Tor Community Gathering in Denmark. The event takes place from 2026-03-13 to 2026-03-15. For more information, check out our website at https://onionize.space/2026/ Thank you all for all the work you do to make Tor better! <3 Cheers, Alex [1]: https://onionize.space/2025/ [2]: https://gitlab.torproject.org/tpo/network-health/margot/ [3]: https://github.com/libressl/portable/blob/a989b7acb9a475fde656e48dbcb38289d… [4]: https://github.com/RustCrypto/rustls-rustcrypto/pull/143 -- Alexander Hansen Færøy

1 0

big german provider "Deutsche Telekom" has started blocking access to their SIP servers
by iji＠gmx.net 25 Feb '26

25 Feb '26

Hi a big German provider "Deutsche Telekom" has started blocking access (at TCP layer) to their SIP servers in February 2026. That means running a middle tor on this provider line results in a non-functional provider SIP service. You have to choose between running a middle tor or a working provider SIP service.

8 18

Happy Families migration status dashboard
by Tor at 1AEO 24 Feb '26

24 Feb '26

To help answer “what still needs to happen before the network can rely on Happy Families (and eventually stop depending on MyFamily)?”, we put together a small public dashboard that tracks the main rollout gates: https://metrics.1aeo.com/network-health.html#happy-families High-level flow: 1) Voting directory authorities upgrade (v0.4.9.x+) DAs need to be on v0.4.9.x+ so the network can move forward with the newer directory protocol pieces. 2) Consensus method moves to 35 Once a >2/3 voting DA majority supports method 35 (7/9), the consensus can adopt it. 3) Relays publish Happy Families material Relay families configure FamilyId so relays publish family-cert, which lets directory documents/micro descriptors carry the happy families identifier information. 4) Clients upgrade and use it As Tor clients (Tor Browser, tor, etc.) widely consume happy families identifiers, then deprecating MyFamily becomes possible. The dashboard shows live counts for DA readiness (v0.4.9.x+ and method-35 support), current consensus method, relay publication (family-cert), and a few “how much of the network is ready” slices (guards/exits/CW/bandwidth). References: - Relay operator guide (FamilyId / Happy Families setup): https://community.torproject.org/relay/setup/post-install/family-ids/ - Proposal 321: https://spec.torproject.org/proposals/321-happy-families.html Feedback welcome—especially if there are other readiness signals you’d like to track, or if you spot anything that looks off for your relays. On Tuesday, February 3rd, 2026 at 5:00 AM, Alexander Hansen Færøy via tor-relays <tor-relays(a)lists.torproject.org> wrote: > On 02/02/2026 22.36, tor_appliedprivacy.net via tor-relays wrote: > > Today we enabled the happy family setup across all our relays. > > > > Since all our relays run in a single /24 we also removed MyFamily at the > > same time. This should help reduce the overhead introduced by MyFamily. > > This is not what any of us on the Network Team are recommending right now, but > if others think this is a good idea despite that: please ensure that your IPv6 > addresses, at the very least, are also within the same /32 in addition to the > /16 requirement for IPv4. I am almost certain that is always the case, but > please just be sure before you proceed. > > We are very interested in hearing feedback on the 0.4.9.4-rc release candidate > and in resolving any issues before we stabilize 0.4.9. Maybe the above approach > will reveal some issues :-) > > Cheers, > Alex > > -- > Alexander Hansen Færøy > _______________________________________________ > tor-relays mailing list -- tor-relays(a)lists.torproject.org > To unsubscribe send an email to tor-relays-leave(a)lists.torproject.org >

1 0

Question about Tor relay guard/middle probability fluctuations
by Tor-relay 23 Feb '26

23 Feb '26

Hi everyone, I have a question regarding my Tor relay’s lifecycle. Since the beginning, I have noticed that the guard and middle probabilities fluctuate significantly. I have not observed this behavior when comparing with other operators’ relays. After some time, the relay receives the Guard flag and the middle probability decreases, later it loses the Guard flag and the middle probability increases again. I waited for a long time to see whether this would stabilize, but the behavior has remained unchanged. I would like to ask what the possible reasons for this could be, and whether it might be related to my configuration. Relay fingerprint: 8E6E3B11C9AD2E9E3B3FC53A4E43D1AE3BEBE444 Tor metrics link: https://metrics.torproject.org/rs.html#details/8E6E3B11C9AD2E9E3B3FC53A4E43… Thank you in advance. P.S. If you check the relay statistics, please ignore the low consensus weight and probabilities during the last month, that was caused by my own mistake.

3 3

the case for relay state persistency
by Gilberto 23 Feb '26

23 Feb '26

Hello everyone, In our diskless infrastructure [0][1], we store all configurations generated by the node inside the TPM's memory. These memories are very limited (1-9k), but more than sufficient for saving the long-term keys of the relays. However, we have a problem: when rebooting the machines, we lose the state file and, with it, the bandwidth measurements of the various relays. This is quite annoying because upon reboot, our relays all start up announcing 0 B/s, and it takes them days to regain their bandwidth. Unfortunately, these files are too large to be stored in the TPM. We would like to be able to generate the BWHistory configurations based on some arbitrary values that we write into the node configurations, and then let the node handle redoing the measurements upon reboot. However, among all the configurations, it's not clear to us which ones are important for our needs and which others we can leave empty. In general, we would like to open a discussion, in view of Arti, about the organization of the datadir. Currently, it's not very clear which files are important for runtime, which are for persistence, etc. From our point of view, it would be quite useful to be able to group persistent data into just a few files, for example, one for all keys and one for states/configurations. Thank you, g [0]: https://osservatorionessuno.org/blog/2025/05/patela-a-basement-full-of-amne… [1]: https://osservatorionessuno.org/blog/2025/12/patela-v2-from-certificates-to…

4 4

Re: Issues with My Tor Relay
by stay.hydrated834＠passmail.net 22 Feb '26

22 Feb '26

Hi everyone, Thank you for the reply and the information you provided. While preparing my response about the blocked traffic, I wanted to clarify the NAT and WAN rules in the screenshots and replace the aliases with the actual IP addresses. During that process I discovered that I (the numb nut of an administrator) forgot to fill in the NAT alias, so there was never an active NAT rule only the WAN rule was present. Because both the NAT and WAN rules are now active, the Tor relay is functioning correctly: no traffic to port 9001 is being blocked, and outbound traffic is unrestricted. I wonder why some connections were allowed by the WAN rule while others were denied and are now only allowed due to the NAT rule. Bandwidth: Average throughput: ~60 Mbps (occasionally up to 140 Mbps) Recent spikes: short bursts up to 420 Mbps training i guess? Due to frequent firewall restarts while I’m re‑configuring my network. I expect the traffic to stabilize as the network settles and after the training finished, and I’m happy if the relay can make use of any unused capacity. 1000Mbps would be the Maximum so there is alot of room to grow. :D Unexpected traffic bursts Every few hours I see dozens of Tor nodes connecting to my relay for 2–5 minutes, resulting in 1,000–7,500 connections dropping. These bursts don’t line up with the bandwidth spikes as far as i seen. My CPU usage, which is usually around 1 %, can rise to about 1.2 % during those periods. Im glad i already thought its static. :D I don’t think this is an attack—nothing I’ve seen, even a smart bulb, generates comparable traffic. I’m wondering if I still have a misconfiguration somewhere. Regarding Snowflakes Is it possible to Route all of my phones (apps) traffic still through WireGuard back to home while allowing Orbot to use the mobile network to open a Snowflake relay (my IP changes every 24 hours after a phone reboot). I have an unlimited 25Mbit not fast but steady. Im on GrapheneOS Screenshots I’ve attached screenshots that highlight exactly what is being blocked, so you can see the traffic patterns I described. Thanks again to everyone who runs relays; your experience helped me finally get mine working. <3 Stay hydrated! Cheers

1 0

Scheduled migration: Nothing to hide infrastructure (23-02-2026)
by Nothing to hide 20 Feb '26

20 Feb '26

Hi everyone, This is a heads-up that a significant portion of the Nothing to hide infrastructure will be undergoing a major upgrade and migration on Monday 23 February 2026. We are replacing older servers and networking equipment with modern high-performance hardware and moving to a new data center. What to expect: * While we aim to limit downtime to a few minutes, extended outages are possible depending on how the migration proceeds. * In the days following the migration, many of our existing relays will be assigned new IP addresses and/or roles. * Once we consider the new network and infrastructure stable, we will be adding more relays to the network. This move also provides much-needed headroom for auxiliary services we plan to offer Tor operators, such as DoT, DoH, DoQ, and NTS. We'll share more details on these in the near future. Best regards, Nothing to hide

1 0

Programmatically determine if a relay is hibernating
by forest-relay-contact＠cryptolab.net 18 Feb '26

18 Feb '26

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Hello. What is the authoritative way to determine if a relay is hibernating? I would have thought that using the control socket would be best, but Tor seems to unlink it when it enters hibernation, so Stem does not work. There's a possible heuristic by checking if /run/tor/tor.pid exists when /run/tor/control does not, but that's fragile because it assumes that the control socket is even enabled. Parsing the logs is doable, but also quite fragile. And I could certainly parse /var/lib/tor/state, but its format is surely subject to change at any time. How do I correctly determine whether a Tor process is in hibernation? Regards, forest -----BEGIN PGP SIGNATURE----- iHUEARYKAB0WIQQtr8ZXhq/o01Qf/pow+TRLM+X4xgUCaZERAgAKCRAw+TRLM+X4 xshIAQCjNVgq0vBzT+MJhbn8Vetj3UObU8DinVvp3gtj9DeLKgD/aKy7N3nvCUSu zFSSFZTCR0MFxQh9TxjmYO/tKRrPoAE= =gAnQ -----END PGP SIGNATURE-----

3 3

handle malicous IPv6 systems abusing the /64 hostmask
by Toralf Förster 15 Feb '26

15 Feb '26

I extended my DDoS solution [1] by blocking systems where malicous connections attempts where observed from up to 256 IPv6 addresses of the same /64 block within 24 hours from hosters providing /64 hostmasks. For now it is rather a quirk than a generic solution. But it works well. And it extends the solution to accept now manual block requests like ipset add tor-ddos6-5443 <IPv6 address> -exist [1] https://github.com/toralf/torutils?tab=readme-ov-file#details -- Toralf

2 2

Tor Node infected with ransomware
by skankhunt42 14 Feb '26

14 Feb '26

Hello there, today I woke up to an execution error of the relayor playbook. I then tried to look into the affected node (tor-nl1.skankhunt42.pw; nickname skankhunt42nl1) and couldn't SSH into it. So I went to the hosters VNC console and found a ransomware notice: Your files are encrypted, requires payment for decrypting Contact us: Telegram: @cloudcone_raidbot UUID: bfaa20d9-7b11-417d-a702-cfa95d6c203c I then tried to boot into recovery and look at the disk but as expected, partition table and ext4 superblocks were gone. hexdump head of the disk was just the ransomware note shown above. I was running Ubuntu 24.04 Minimal with ESM enabled and unattended-upgrades, everything else managed by relayor. I obviously checked the other nodes for unsual SSH logins (as they had the same SSH key) and didn't found anything. I am rotating the keys for now and shut down the VPS at HostSlick. Not sure if there is something to further investigate maybe. What's odd is that I couldn't find anything about "cloudcone_raidbot", doesn't even exist on telegram. I really want to understand what I did wrong. Maybe someone with more experience may take a look at it? Best, skankhunt42

4 4