[tor-talk] single entity running >36% exit probability with 5 relays? scary.

David Goulet dgoulet at ev0ke.net
Thu Dec 10 19:25:01 UTC 2015

On 10 Dec (19:04:02), nusenu wrote:
> take a look at the "e_prob" column:
> https://github.com/nusenu/tor-network-observations/blob/master/2015-12-10_relay-families.txt
> How did he manage to get such a high consensus weight
> (~10x ipredator's cw)?
> these four relays were first seen on 2015-12-10 04:00:00
> are down since a few hours now
> https://atlas.torproject.org/#search/contact:robgjansen


This morning I noticed a huge spike in the estimated exit probability
per continent graph and network capacity graph:


After a bit of investiguation turns out 3 new relays were advertising a
HUGE bandwidth.

I contacted the person (it's a known Tor researcher at NRL) and he
realized they were mis-configured advertising way too much bw. It should
be fixed soon hopefully but they are 3 real exit nodes.

However, the humongeous weights are concerning since they were very
"young" relays in the lifetime of a relay so it worth investiguating and
thinking if this is normal or maybe a bug?


> btw: thanks for setting an apparently proper myfamily config.
> -- 
> tor-talk mailing list - tor-talk at lists.torproject.org
> To unsubscribe or change other settings go to
> https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-talk
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 603 bytes
Desc: Digital signature
URL: <http://lists.torproject.org/pipermail/tor-talk/attachments/20151210/83f2259e/attachment.sig>

More information about the tor-talk mailing list