New relay in USA: bridge or middle relay?
Good morning from Wisconsin, After reading about how middle relays in the USA go largely underutilized, and having quietly run my own middle relay for several years, would it be more beneficial to the network to launch several new bridges instead of more middle relays? Make your day great, Isaac Grover, Senior I.T. Consultant Aileron I.T. - " Because #ProactiveIsBetter " O: 715-377-0440, F:715-690-1029, W: www.aileronit.com
Hi,
On 8 Oct 2019, at 22:07, Isaac Grover, Aileron I.T. <igrover@aileronit.com> wrote:
Good morning from Wisconsin,
After reading about how middle relays in the USA go largely underutilized, and having quietly run my own middle relay for several years, would it be more beneficial to the network to launch several new bridges instead of more middle relays?
Good question. Very few tor relays are actually under-utilised. Many operators expect 100% utilisation, but low-latency protocols work best around 10% utilisation. We're currently at 30%. So feel free to deploy a middle, and if it's fast and stable enough, it might become a guard. Some bridges are kept in reserve. Others are handed out using less popular methods. So feel free to deploy multiple bridges on the same IP address or subnet. T -- teor ----------------------------------------------------------------------
hah, mine is -=severely=- under utilized...... NO CPU Load: https://puu.sh/EyX6N/81a5d5c76e.png 4 Mbps of throughput 2 Mbps each way or only ~ 20Mbps ea way: https://puu.sh/EyX7F/b7885ce635.png Plenty of bandwidth: https://puu.sh/EyX9O/65334af451.png Even to Germany: https://puu.sh/EyXbI/33cb46ac55.png Even to Brazil: https://www.speedtest.net/result/8719378576 Even to London: https://puu.sh/EyXk4/f86e74e856.png I realize that "false advertising of bandwidth to abuse the network protocols" has impacted the "consensus weight" assigned to various nodes. But there -definitely- needs to be a more intelligent system developed for determining this. I just proved that I have 2 GIGABITS of bandwidth.... HUNDREDS to other countries, but there is 20 MegaBits flowing through my relay, lol. Thanks, Matt Westfall President & CIO ECAN Solutions, Inc. Everything Computers and Networks 804.592.1672 ------ Original Message ------ From: "teor" <teor@riseup.net> To: tor-relays@lists.torproject.org Sent: 10/22/2019 4:53:25 AM Subject: Re: [tor-relays] New relay in USA: bridge or middle relay?
Hi,
On 8 Oct 2019, at 22:07, Isaac Grover, Aileron I.T. <igrover@aileronit.com> wrote:
Good morning from Wisconsin,
After reading about how middle relays in the USA go largely underutilized, and having quietly run my own middle relay for several years, would it be more beneficial to the network to launch several new bridges instead of more middle relays?
Good question.
Very few tor relays are actually under-utilised.
Many operators expect 100% utilisation, but low-latency protocols work best around 10% utilisation. We're currently at 30%.
So feel free to deploy a middle, and if it's fast and stable enough, it might become a guard.
Some bridges are kept in reserve. Others are handed out using less popular methods. So feel free to deploy multiple bridges on the same IP address or subnet.
T
-- teor ----------------------------------------------------------------------
Hi,
On 31 Oct 2019, at 10:04, ECAN - Matt Westfall <mwestfall@ecansol.com> wrote:
hah, mine is -=severely=- under utilized......
NO CPU Load: https://puu.sh/EyX6N/81a5d5c76e.png
4 Mbps of throughput 2 Mbps each way or only ~ 20Mbps ea way: https://puu.sh/EyX7F/b7885ce635.png
Plenty of bandwidth: https://puu.sh/EyX9O/65334af451.png
...
I realize that "false advertising of bandwidth to abuse the network protocols" has impacted the "consensus weight" assigned to various nodes.
But there -definitely- needs to be a more intelligent system developed for determining this.
I just proved that I have 2 GIGABITS of bandwidth.... HUNDREDS to other countries, but there is 20 MegaBits flowing through my relay, lol.
The speed tests you ran are not Tor traffic. It's not even clear if they are TCP or TLS. Tor users run Tor clients, just like the bandwidth authorities do. Here are 5 speed tests with actual Tor clients to your relay: https://consensus-health.torproject.org/consensus-health-2019-10-30-23-00.ht... Comcast is well-known for bad peering, and slowing down particular protocols. Search the list archives for details. Here are some steps you can take to try to improve the speed of your relay: https://trac.torproject.org/projects/tor/wiki/doc/MyRelayIsSlow Let us know how you go! T -- teor ----------------------------------------------------------------------
participants (3)
-
ECAN - Matt Westfall -
Isaac Grover, Aileron I.T. -
teor