Re: Tor memory usage on embedded systems.
On Thu, March 5, 2009 20:27, basile wrote: These preliminary numbers might be of interest. Nice work I'm going to repeat these measurements, but would like some feedback from the community regarding what you'd like to see. Could you run the tests after settings the same BandwidthRate and BandwidthBurst for all nodes? I think that a lower rate/burst node should be less used then an higher one. ciao -- Marco Bonetti BT3 EeePC enhancing module: http://sid77.slackware.it/bt3/ Slackintosh Linux Project Developer: http://workaround.ch/ Linux-live for powerpc: http://workaround.ch/pub/rsync/mb/linux-live/ My webstuff: http://sidbox.homelinux.org/ My GnuPG key id: 0x86A91047
Re: Tor memory usage on embedded systems.
pho...@rootme.org wrote:
On Fri, Mar 06, 2009 at 01:12:51AM +0100, sl...@slush.cz wrote 3.6K bytes in
99 lines about:
: Thanks for pointing that out. I'm trying to answer the question what is
: the minimum amount of RAM required to run a bare minimum linux system
: which can support a tor relay/exit/directory node. Suggestions?
The command pmap may also work,
http://linuxcommand.org/man_pages/pmap1.html
It gives you a handy total at the end of its output.
Alternatively, just parse /proc/{tor pid}/status for the details.
Hi Andrew, this is one approach, but I want a system total, not just the
memory usage on a process by process basis. It would be nice to be able
to answer questions like if we want to run a tor exit and directory
server at such and such a rate on an embedded device, how much ram does
the device need?. Tor needs some minimal OS in which to live. The
least I could do is busybox + openntp + tor. The memory requirements of
these processes must be added in. Also embedded devices run purely in
RAM, so the filesystem contributes to usage and tor needs about 30MB in
its DataDirectory. This also needs to be added in. Rather than
identifying all the pieces and adding, which is not the easiest thing to
do without missing something or double counting, the approach I took is
to just ask the system for a total with free. (Eg. pmap needs careful
interpretaion when adding up totals for more than one process because of
shared memory.)
I think my MIPS numbers are good, but my i686 are misleading. slush's
response jarred me to look at how free reports memory usage for
transitional ramdisks (/dev/ramX) devices versus what it does with
initramfs.
--
Anthony G. Basile, Ph.D.
Chair of Information Technology
D'Youville College
Buffalo, NY 14201
USA
(716) 829-8197
signature.asc
Description: OpenPGP digital signature
Re: Tor memory usage on embedded systems.
Marco Bonetti wrote: Could you run the tests after settings the same BandwidthRate and BandwidthBurst for all nodes? I think that a lower rate/burst node should be less used then an higher one. Yes. I'm hoping in the long run to produce something like RAM requirements on embedded systems = function of ( tor services provided, BandwidthRate, other relavent parameters ) where the other parameters may include architecture, uclibc vs glibc etc. Not an exhaustive study, but something of a guide to the community should people want to start putting tor servers on embedded devices along the lines of what JanusPA does. -- Anthony G. Basile, Ph.D. Chair of Information Technology D'Youville College Buffalo, NY 14201 USA (716) 829-8197 signature.asc Description: OpenPGP digital signature
Re: MapAddress not working
On Sun, 1 Mar 2009 20:32:26 -0500, misc wrote: The function MapAddress in tor config no longer works. Tor simply ignores it and connect using any node. Any solutions? Tor 0.2.0.33/Vidalia 0.1.9 What am I doing wrong? Please advise how to make MapAddress setting to work in Tor config! There are forums where I have to constantly re-login because of IP changes. I need to use one IP. It worked before, with older Tor versions. HELP!
Re: MIT Circumvention Landscape Report
Hello Dear Andrew and TOR dev teams,, I can only admire your work, it is a bit over my head to try to help you and is just comprehendable for me to understand much of your efforts. Thanks, Algenon --- On Thu, 3/5/09, pho...@rootme.org pho...@rootme.org wrote: From: pho...@rootme.org pho...@rootme.org Subject: Re: MIT Circumvention Landscape Report To: or-talk@freehaven.net Date: Thursday, March 5, 2009, 11:37 AM -Inline Attachment Follows- On Thu, Mar 05, 2009 at 11:56:17AM -0500, bas...@opensource.dyc.edu wrote 1..1K bytes in 45 lines about: : http://tech.slashdot.org/article.pl?sid=09/03/05/1334220from=rss : : http://cyber.law.harvard.edu/publications/2009/2007_Circumvention_Landscape_Report It's actually Harvard/Berkman, not MIT. And Tor has a response, https://blog.torproject.org/blog/berkman-2007-circumvention-landscape-and-progress Thanks! -- Andrew
