[tor-relays] 3-hourly pattern in network load of my Tor exit relay

Toralf Förster toralf.foerster at gmx.de
Sun Jul 29 10:57:45 UTC 2018

I do run 2 exits at the same hardware [1] and do use systat [2] to monitor system data.

2 questions:

	Since a while (months) I do observe a periodic pattern in the network load. An example is the graph (SVG format, made with [3]) of the data from yesterday.
	Is this typical? I did not observe such a pattern in the last few years.

	htop shows me, that the Tor exit relay listening on ORPort 443 consumes typically 30% more CPU time than the second Tor exit relay listing on ORport 9001. Boths however have nearly identical configs, bandwith weights and amount of connections (typical values are 4300 and 4500 respectively, 3700 and 3900 of them are connections to relays, the delta are exit connections).
	Does this explains the delta in CPU time?

[1] https://metrics.torproject.org/rs.html#search/family:D11D11877769B9E617537B4B46BFB92B443DE33D
[2] http://sebastien.godard.pagesperso-orange.fr/
[3] svg=/tmp/graph.svg; sadf -g -T ~/tmp/sa/sa28 -O skipempty,oneday -- -n ALL > $svg && konqueror $svg


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 228 bytes
Desc: OpenPGP digital signature
URL: <http://lists.torproject.org/pipermail/tor-relays/attachments/20180729/2b80e491/attachment.sig>

More information about the tor-relays mailing list