[tor-relays] Why is my relay dying?

Roman Mamedov rm at romanrm.net
Sat Nov 3 19:01:34 UTC 2018

On Sat, 3 Nov 2018 13:26:17 -0500
Scott Ashcraft <scott at ashcraft.com> wrote:

> https://metrics.torproject.org/rs.html#details/EC0ABA811E4EB33DAD8BC8B7037D862BF4F3AA28
> I honestly have no idea. I am running on Google Fiber and have set:
> RelayBandwidthRate 52000 KBytes
> RelayBandwidthBurst 104000 KBytes
> The spike in January is when I switched out some bad equipment. There were
> a couple of outages in late Feb/Mar for a few hours each, and then some
> (usually monthly) restarts since. I also at one point tried to add a Family
> Member which I have since removed. Other than that, there haven't been any
> changes to my setup. Aside from the addition, then removal of the family
> member, my torrc configuration hasn't changed since starting over a year
> ago.
> Any thoughts? I'd like to be more useful to the community,

What you can do right off the bat, is to run a second Tor instance on the same
IP address (of course on a different port). You can run two per IP, and it is
most often a no-brainer to do so. I would expect it to get around the same 4-5
MB/sec usage over time -- with little to no impact on speeds of the first one.

As for why the bandwidth drops off, maybe you just don't get a lot of download
or upload speed, despite what your plan advertised (it's usually "up to"
anyways). Or the ISP shapes weird ports and protocols, or maybe even Tor, to a
lower speed specifically. Try choosing some standard and commonly used for
"encrypted data" port for the 2nd instance, such as 443/993/995/etc.

Also if you find that you don't reach the RelayBandwidthRate/Burst by a long
shot, it's better to just comment them out entirely, I think it saves some CPU
time on useless housekeeping.

With respect,

More information about the tor-relays mailing list