I'm running a relay with the extra bandwidth on my VPS, but I'm unsure if it's optimally configured.
I have 2.5TB of bandwidth I want to use every month. Given that, how do I configure my relay in the most helpful way?
The docs mention dividing my bandwidth by 30 and setting an accounting max of 1 day. It also suggests "You might also consider rate limiting to spread your usefulness over more of the day". What not use an accouting period of a month and spread it out over the full month? Spreading it out feels like the right thing to do, but is it actually the most helpful? Maybe burning my bandwidth is faster bursts is more helpful.
If spreading it over the month using a low bandwidth rate and a monthly accounting period is ideal, is there a better way to configure it than busting out a calculator and coverting TB/month to KB/s?
How should I determine my BandwidthBurstRate? I understand what it is, but should it be defined in relation to my bandwidth rate?
Thanks in advance, Curtis Gagliardi
On Wed, 29 Apr 2015 17:44:38 -0700, Curtis Gagliardi curtis@curtis.io wrote:
I'm running a relay with the extra bandwidth on my VPS, but I'm unsure if it's optimally configured.
I have 2.5TB of bandwidth I want to use every month. Given that, how do I configure my relay in the most helpful way?
The docs mention dividing my bandwidth by 30 and setting an accounting max of 1 day. It also suggests "You might also consider rate limiting to spread your usefulness over more of the day". What not use an accouting period of a month and spread it out over the full month? Spreading it out feels like the right thing to do, but is it actually the most helpful? Maybe burning my bandwidth is faster bursts is more helpful.
If spreading it over the month using a low bandwidth rate and a monthly accounting period is ideal, is there a better way to configure it than busting out a calculator and coverting TB/month to KB/s?
How should I determine my BandwidthBurstRate? I understand what it is, but should it be defined in relation to my bandwidth rate?
I'm running an exit node on an entry level VULTR VPS which comes with 1000GB per month.
This is all I put in the torrc file to limit bandwidth usage:
# Bandwidth and data caps AccountingStart day 19:45 # calculate once a day at 7:45pm AccountingMax 33 GBytes
It's surprisingly accurate. As of today VPS usage is at 98% of monthly allowance. I didn't bother with burst rate because it never seems to climb over 10Mbps usage, and the interface is a 100Mbps connection.
As far as letting in run full steam and then shut down for potentially hours every day vs. finding a steady rate that it can burn bandwidth at for the entire month, that's been discussed already in the list archives here: http://www.mail-archive.com/tor-relays%40lists.torproject.org/msg05478.html
On Wed, 29 Apr 2015 17:44:38 -0700, Curtis Gagliardi curtis@curtis.io wrote:
Spreading it out feels like the right thing to do, but is it actually the most helpful? Maybe burning my bandwidth is faster bursts is more helpful.
Also found these relevant threads by searching the archives for 'hibernate' :
http://www.mail-archive.com/tor-relays%40lists.torproject.org/msg05002.html http://www.mail-archive.com/tor-relays@lists.torproject.org/msg05569.html
There was some other discussion (can't remember exactly where off-hand, might have been another lits) disputing the Tor manual's stance of "better have a fast relay part of the time than a slow relay all the time". Maybe this is case dependent on what the minimum bandwidth levels are, dunno.
tor-relays@lists.torproject.org