Hello,
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center and actual customers have not complained about any performance issues.
Operating systems and Tor version are up to date. I'm dedicating a significant portion of bandwidth to these nodes - 10gbit/sec.
Am I having issues with the bandwidth authorities?
I'm growing frustrated with my performance to resources ratio, I should be doing far better than this.
Please throw ideas at me - open to any ideas.
Thanks! John Quintex Alliance Consulting
On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
It seems disproportionate though. I'm only using 2gbit/sec of my circuit and I have plenty of hardware ceiling left. Feels like i'm doing something else wrong.
On Jan 7, 2020, at 07:30, Toralf Förster toralf.foerster@gmx.de wrote:
On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
-- Toralf
tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
Months ago John beat my exit consensus weight. The Tor network didn’t grew substantially in this timeframe but he lost around 2/3 of his traffic while I was stable.
On 7. Jan 2020, at 14:30, Toralf Förster toralf.foerster@gmx.de wrote:
Signed PGP part On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
-- Toralf
I also would like to add to this - if it were just the Tor network increasing in size I could see my consensus weight dropping and my bandwidth staying the same. I'm simply not getting the 7-8gbit/sec traffic I was. Truly odd.
-----Original Message----- From: tor-relays tor-relays-bounces@lists.torproject.org On Behalf Of Toralf Förster Sent: Tuesday, January 7, 2020 7:30 AM To: tor-relays@lists.torproject.org Subject: Re: [tor-relays] Consensus Weight Dropping/Authority Issues?
On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
Consensus & usage are independent consensus: based on available bandwidth load: based on usage by tor clients.
if total available bw increases but load doesn't, observed load on a node will drop.
On Tue, 7 Jan 2020 at 17:27, John Ricketts john@quintex.com wrote:
I also would like to add to this - if it were just the Tor network increasing in size I could see my consensus weight dropping and my bandwidth staying the same. I'm simply not getting the 7-8gbit/sec traffic I was. Truly odd.
-----Original Message----- From: tor-relays tor-relays-bounces@lists.torproject.org On Behalf Of Toralf Förster Sent: Tuesday, January 7, 2020 7:30 AM To: tor-relays@lists.torproject.org Subject: Re: [tor-relays] Consensus Weight Dropping/Authority Issues?
On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50
exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
-- Toralf
tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
On 7. Jan 2020, at 18:20, r1610091651 r1610091651@telenet.be wrote:
Consensus & usage are independent consensus: based on available bandwidth load: based on usage by tor clients.
if total available bw increases but load doesn't, observed load on a node will drop.
We are talking about Exists.
b n
On Tue, 7 Jan 2020 at 17:27, John Ricketts <john@quintex.com mailto:john@quintex.com> wrote: I also would like to add to this - if it were just the Tor network increasing in size I could see my consensus weight dropping and my bandwidth staying the same. I'm simply not getting the 7-8gbit/sec traffic I was. Truly odd.
-----Original Message----- From: tor-relays <tor-relays-bounces@lists.torproject.org mailto:tor-relays-bounces@lists.torproject.org> On Behalf Of Toralf Förster Sent: Tuesday, January 7, 2020 7:30 AM To: tor-relays@lists.torproject.org mailto:tor-relays@lists.torproject.org Subject: Re: [tor-relays] Consensus Weight Dropping/Authority Issues?
On 1/7/20 1:57 PM, John Ricketts wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center
Which correlates to https://metrics.torproject.org/bandwidth.html https://metrics.torproject.org/bandwidth.html - your fraction just decreases if more and more relays join the party.
-- Toralf
tor-relays mailing list tor-relays@lists.torproject.org mailto:tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
consensus means what fraction of traffic will pass over your nodes, statistically speaking. Hence a steady drop of consensus value, with no infra changes on your end, could also be explained by a stead rise of total bandwidth available: since your part is fixed and total grows, your fraction reduces.
Regards
On Tue, 7 Jan 2020 at 14:04, John Ricketts john@quintex.com wrote:
Hello,
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center and actual customers have not complained about any performance issues.
Operating systems and Tor version are up to date. I'm dedicating a significant portion of bandwidth to these nodes - 10gbit/sec.
Am I having issues with the bandwidth authorities?
I'm growing frustrated with my performance to resources ratio, I should be doing far better than this.
Please throw ideas at me - open to any ideas.
Thanks! John Quintex Alliance Consulting _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
Totally agree with your analysis on this- my concern is that if that were true I'd see Nifty's 15 percent drop like a rock too... and drop from 7-8gbit/sec to 2-3gb/sec is weird.
You'd think that I'd be saturated. I've been running these nodes approximately three years and I've never seen this.
John
On Jan 7, 2020, at 11:28, r1610091651 r1610091651@telenet.be wrote:
? consensus means what fraction of traffic will pass over your nodes, statistically speaking. Hence a steady drop of consensus value, with no infra changes on your end, could also be explained by a stead rise of total bandwidth available: since your part is fixed and total grows, your fraction reduces.
Regards
On Tue, 7 Jan 2020 at 14:04, John Ricketts <john@quintex.commailto:john@quintex.com> wrote: Hello,
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center and actual customers have not complained about any performance issues.
Operating systems and Tor version are up to date. I'm dedicating a significant portion of bandwidth to these nodes - 10gbit/sec.
Am I having issues with the bandwidth authorities?
I'm growing frustrated with my performance to resources ratio, I should be doing far better than this.
Please throw ideas at me - open to any ideas.
Thanks! John Quintex Alliance Consulting _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.orgmailto:tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
Do you have any insights in your DNS performance (response time, cache hitrate) and qps rate over time?
(please don't publish it)
Hi John,
On 7 Jan 2020, at 22:57, John Ricketts john@quintex.com wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center and actual customers have not complained about any performance issues.
Operating systems and Tor version are up to date. I'm dedicating a significant portion of bandwidth to these nodes - 10gbit/sec.
Am I having issues with the bandwidth authorities?
I'm growing frustrated with my performance to resources ratio, I should be doing far better than this.
Did you ever find an answer here?
What have you analysed? Have you tried any config changes?
Can you tell us which directory authorities are measuring your relays lower than they were before?
The most likely scenarios are: * Routing changes between your relays and the bandwidth authorities * The Torflow to sbws transition * Did you upgrade your tor version? Most of the network upgraded to tor 0.4.1 and 0.4.2 recently: https://metrics.torproject.org/versions.html?start=2019-09-01&end=2020-0...
Did the consensus weight drop first, or did the observed bandwidth drop first?
You've probably read this wiki page before, but just in case: https://trac.torproject.org/projects/tor/wiki/doc/MyRelayIsSlow#FindingOutwh...
T
Hi all,
On 16. Mar 2020, at 07:43, teor teor@riseup.net wrote:
On 7 Jan 2020, at 22:57, John Ricketts john@quintex.com wrote:
I have been watching the consensus weight and bandwidth of all of my 50 exit nodes drop consistently over the past few months. I have not made any hardware changes in my data center and actual customers have not complained about any performance issues.
Operating systems and Tor version are up to date. I'm dedicating a significant portion of bandwidth to these nodes - 10gbit/sec.
Am I having issues with the bandwidth authorities?
I'm growing frustrated with my performance to resources ratio, I should be doing far better than this.
Did you ever find an answer here?
What have you analysed? Have you tried any config changes?
Can you tell us which directory authorities are measuring your relays lower than they were before?
The most likely scenarios are:
- Routing changes between your relays and the bandwidth authorities
- The Torflow to sbws transition
- Did you upgrade your tor version?
Most of the network upgraded to tor 0.4.1 and 0.4.2 recently: https://metrics.torproject.org/versions.html?start=2019-09-01&end=2020-0...
Did the consensus weight drop first, or did the observed bandwidth drop first?
You've probably read this wiki page before, but just in case: https://trac.torproject.org/projects/tor/wiki/doc/MyRelayIsSlow#FindingOutwh...
T
Also I want to take this opportunity to say I'm desperately trying to resurrect the old-style bw scanner on gabelmoo, but it isn't going too well. Sorry if this outage is causing any kind of issues for anyone :(
Cheers Sebastian
tor-relays@lists.torproject.org
-
John Ricketts -
niftybunny -
nusenu -
r1610091651 -
Sebastian Hahn -
teor -
Toralf Förster