[To the maintainers of Tor Metrics] Metrics showing erroneous data for my two relays, can somebody responsible please take a look at the issue?
Hello,
for about 3 months I have been hosting two Tor relays:
https://metrics.torproject.org/rs.html#details/AF42E6C77196A37F041A1A1E953E5...
https://metrics.torproject.org/rs.html#details/7F7D1A5BE88FA7C9358955705AE7A...
After doing system maintenance (mainly upgrading packages), slightly adjusting my "BandwidthRate" and "BandwidthBurst" for both relays and then rebooting the server earlier, the guard relay now shows the following on Metrics, even though the graphs still show the old, correct data:
The first time that this relay was seen in the consensus: 2024-01-10 16:00:00 (8 hours 57 minutes and 21 seconds)
This is even though metrics shows an uptime of almost 16 hours?!
The fingerprints did not change according to both the metrics page, and the Tor log.
I did not delete or mess with each relays DataDirectory or the secret keys.
I have a MyFamily setting in both relays configs:
MyFamily $AF42E6C77196A37F041A1A1E953E51B4656BDC1B, $7F7D1A5BE88FA7C9358955705AE7AFA61EEDA2B0
Yet the relays don't show up as being in a family, and Metrics doesn't at all display the other relay counterpart as a family member when visiting the page for the guard or exit.
There is also no line in the relays journal (from journalctl -u tor@guard) indicating that the guard relay is new to the network, and it retains its older flags.
This only seems to affect my guard relay, the exit metrics page shows mostly valid data, except for the missing family member.
Since I did not change anything that would cause this, this is likely a bug with the metrics page, and I would encourage you to look into it.
Thank you very much, George
Small update:
Both metrics pages now show the correct family fingerprints, but the first seen date for the guard node is still wrong.
It should be:
2023-11-26 05:00:00 (46 days 9 hours 10 minutes and 42 seconds)
but it shows:
2024-01-10 16:00:00 (22 hours 14 minutes and 17 seconds)
Everything else seems fixed.
Regards, George On Thursday, January 11th, 2024 at 2:18 AM, George Hartley via tor-relays tor-relays@lists.torproject.org wrote:
Hello,
for about 3 months I have been hosting two Tor relays:
https://metrics.torproject.org/rs.html#details/AF42E6C77196A37F041A1A1E953E5...
https://metrics.torproject.org/rs.html#details/7F7D1A5BE88FA7C9358955705AE7A...
After doing system maintenance (mainly upgrading packages), slightly adjusting my "BandwidthRate" and "BandwidthBurst" for both relays and then rebooting the server earlier, the guard relay now shows the following on Metrics, even though the graphs still show the old, correct data:
The first time that this relay was seen in the consensus: 2024-01-10 16:00:00 (8 hours 57 minutes and 21 seconds)
This is even though metrics shows an uptime of almost 16 hours?!
The fingerprints did not change according to both the metrics page, and the Tor log.
I did not delete or mess with each relays DataDirectory or the secret keys.
I have a MyFamily setting in both relays configs:
MyFamily $AF42E6C77196A37F041A1A1E953E51B4656BDC1B, $7F7D1A5BE88FA7C9358955705AE7AFA61EEDA2B0
Yet the relays don't show up as being in a family, and Metrics doesn't at all display the other relay counterpart as a family member when visiting the page for the guard or exit.
There is also no line in the relays journal (from journalctl -u tor@guard) indicating that the guard relay is new to the network, and it retains its older flags.
This only seems to affect my guard relay, the exit metrics page shows mostly valid data, except for the missing family member.
Since I did not change anything that would cause this, this is likely a bug with the metrics page, and I would encourage you to look into it.
Thank you very much, George
network-health@lists.torproject.org
-
George Hartley