Non-exit relay appears to work, but reachability self-test is failing and descriptor is not being published (tor 0.4.6.9).

I've been running a non-exit relay in a bare-metal debian bullseye machine, behind NAT, for some days and it appears to be "working".

But after starting/restarting tor I got the following:

[NOTICE] Now checking whether IPv4 ORPort <my public ip>:443 is reachable... (this may take up to 20 minutes -- look for log messages indicating success)

And after 20 minutes:

[WARN] Your server has not managed to confirm reachability for its ORPort(s) at <my public ip>:443. Relays do not publish descriptors until their ORPort and DirPort are reachable. Please check your firewalls, ports, address, /etc/hosts file, etc.

And then this message repeats every 20 minutes.

Tor self-reachability test appears to be failing and my relay descriptor appears not to be published.

But the port is open and reachable from outside my lan (I can confirm using nmap, for example) and the relay appears to be "relaying" traffic.

Using nyx to monitor its activity I can see constant connections (outbound, circuit and directory) on the correspondent tab and a considerable amount of traffic.

I cannot find my relay in the metrics site or in the collector site, but I can see it listed in the consensus health site (it appears to have votes from every authority, but the consensus column is blank - also no data for bw).

I also get this when starting/restarting tor:

[notice] External address seen and suggested by a directory authority: <my public ip>

And this with every "Heartbeat" (6h intervals):

[NOTICE] Heartbeat: It seems like we are not in the cached consensus.

The version of tor being used is the one in the stable branch of the Tor Project's apt repository (0.4.6.9).

Relevant lines of my torrc file:

ORPort 443 IPv4Only
ExitRelay   0
ExitPolicy reject *:*
AddressDisableIPv6 1

Edited Feb 08, 2022 by sacou

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information