1

I have two relays that started life about a month ago so we are sill in the "Lifecycle of a new relay" time frame. The differences in the two are physical location and ORPort. Both are on Digital Ocean VPS. 1cpu 1GB memory. Both are Debian 8.6 and tor 0.2.8.11

Relay 1 is located in San Francisco and ORPort port is set to: 9001

Relay 2 is located in Singapore and ORPort is set to 443

Bandwidth on both relays is: 3168 KBytes Rate/Burst

IPv6 is enabled on both.

Relay #1 gets quite a bit more traffic than relay #2. Usually 3-4 times more but sometimes even more according to atlas.torproject.org

I'm not much of a sys admin so I'm not sure how to calculate the data transfers accurately. I use arm and glances to get an idea. arm seems to way off on its 'average' report.

At some point in the future I will have a reason to restart these relays (software upgrades etc) so I'm wondering if the difference is due to location, ORPort or both. Then maybe I can get Singapore to be a little more efficient.

I also have a plan to install a bridge in the future and would possibly choose Singapore for location.

Added info: atlas reports 'Advertised Bandwidth' on relay #1 3.19 MB/s and relay #2 991.26 KB/s

torstatus.blutemagie.de shows

Relay #1 3244032 / 3244032 / 334499 for Rate/Max/Observed in Bps

Relay #2 3244032 / 3244032 / 41903 for Rate/Max/Observed in Bps

  • What does Atlas report as the measured/observed bandwidth for each? It's likely that Singapore has less good peering with the rest of the Tor Network and might suffer because of it. – cacahuatl Jan 9 '17 at 10:29
  • (It's unlikely to be the ORPort, that mostly matters for clients connecting directly to you) – cacahuatl Jan 9 '17 at 10:51
  • See added info in original post. I'm not comfortable with markdown yet. Still learning – jmd8800 Jan 9 '17 at 14:27
  • I'm pretty sure this is going to be an issue with BWAuth's measured values being worse due to peering with Singapore. I've seen similar situations before with the exact same DC. None of the other relays are much more than 1MB/s and the one that is, is in a different DC. – cacahuatl Jan 9 '17 at 17:43
  • Thanks for the response. I'll likely keep it running for geo diversity but trim back the expense as more than 1/2 of the configuration isn't being utilized. – jmd8800 Jan 10 '17 at 2:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.