Hi,
Looking at the graphs showing the number of relays in the network it seems like we've lost about 500 (-7%) relays since the beginning of this year.
https://metrics.torproject.org/networksize.html?graph=networksize&start=... https://metrics.torproject.org/networksize.html?graph=networksize&start=...
Looking at the graphs showing the total relay bandwidth of the network it seems like the advertised bandwidth has increased with about 25 Mbps (+23%).
https://metrics.torproject.org/bandwidth.html?graph=bandwidth&start=2015... https://metrics.torproject.org/bandwidth.html?graph=bandwidth&start=2012...
Seems a bit contradictory at first sight. A guess would be that a lower number of fast relays have replaced a higher number of slow ones but I haven't looked into it more.
Anyone who's looked into this? And can back up their theory with some numbers.
Thanks, Linus
On Sunday, May 3, 2015 10:08am, "Linus Nordberg" linus@nordberg.se said:
Hi,
Looking at the graphs showing the number of relays in the network it seems like we've lost about 500 (-7%) relays since the beginning of this year.
https://metrics.torproject.org/networksize.html?graph=networksize&start=... https://metrics.torproject.org/networksize.html?graph=networksize&start=...
My uninformed guess would be that the higher minimum bandwidth requirements in v0.2.6.x forced out the marginal relays.
"Steve Snyder" swsnyder@snydernet.net wrote Sun, 3 May 2015 10:40:59 -0400 (EDT):
| My uninformed guess would be that the higher minimum bandwidth requirements in v0.2.6.x forced out the marginal relays.
Interesting. I'll see if I can find out when a majority of directory authorities upgraded to 0.2.6.x.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512
Looking at the graphs showing the number of relays in the network it seems like we've lost about 500 (-7%) relays since the beginning of this year.
https://metrics.torproject.org/networksize.html?graph=networksize&start=...
https://metrics.torproject.org/networksize.html?graph=networksize&start=...
Looking at the graphs showing the total relay bandwidth of the network it seems like the advertised bandwidth has increased with about 25 Mbps (+23%).
I doubt that 25 Mpbs is 23% of the tor network capacity, I guess it should say GBit/s.
https://metrics.torproject.org/bandwidth.html?graph=bandwidth&start=2015...
https://metrics.torproject.org/bandwidth.html?graph=bandwidth&start=2012...
Seems a bit contradictory at first sight. A guess would be that a lower number of fast relays have replaced a higher number of slow ones but I haven't looked into it more.
Anyone who's looked into this? And can back up their theory with some numbers.
Unfortunately I've got no historic onionoo details data that goes back to 2014 to compare with current data, but if anyone has, feel free to upload it somewhere.
You probably also know about the 0.2.3.x relays: https://metrics.torproject.org/versions.html
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512
nusenu nusenu@openmailbox.org wrote:
it seems like we've lost about 500 (-7%) relays since the beginning of this year.
the Tor network recently lost a large amount of relays (~75 or more) due to BadExit attacks. this may account for a large sum of the disappearing nodes (~20%), but it doesn't account for nearly all of them.
nusenu nusenu@openmailbox.org wrote Sun, 03 May 2015 19:06:39 +0000:
| > Looking at the graphs showing the total relay bandwidth of the | > network it seems like the advertised bandwidth has increased with | > about 25 Mbps (+23%). | | I doubt that 25 Mpbs is 23% of the tor network capacity, I guess it | should say GBit/s.
Indeed.
| You probably also know about the 0.2.3.x relays: | https://metrics.torproject.org/versions.html
I didn't. Thanks.
tor-relays@lists.torproject.org
-
Linus Nordberg -
nusenu -
Steve Snyder -
syndikal