On Fri, Aug 15, 2025 at 06:29:40AM -0000, tor--- via tor-relays wrote:
ADA11AA7FD14CFE904C7E302CA08EFBC8AFB415D
My relay jumped to 13500 weight, then back to 1. Even thought relay is really young, this is (in my opinnion) not normal behavior?
Tor notice logs doesnt give any error, and relay got unmettered bandwidth.
Thanks for running a relay! You can see the individual bandwidth votes about your relay by going to the bottom of https://consensus-health.torproject.org/#relayinfo and putting in your fingerprint. The directory authorities use the "low median" among the bw values (once there are at least 3) to decide what your consensus weight should be. I see that your consensus weight is back up now, but some of the directory authorities still vote quite low numbers for you. We could go reconstruct each of the individual hourly votes from the data at https://collector.torproject.org/recent/ but I have not done so. Maybe one of the directory authorities that was voting high for you stopped voting for a bit and that's how you dropped back down? My guess is that in the next few days the dir auths that are voting low will measure you against your new higher self-reported bandwidth and then also give you a higher vote, and things will stabilize. (If you weren't in northern Europe I would have a different answer, which is that our measurement approach is over-estimating bandwidth in northern Europe and underestimating it elsewhere, because we measure with two-hop circuits so relays that are near other fast relays will end up seen as faster. This is a problem for overall network load balancing, and for overall security since we concentrate too much traffic in that region, but in your particular case it should help your relay end up with a higher consensus weight once there have been enough measurements.) --Roger