15:00:05 <skalnik1> #startmeeting Power management for Fedora
15:00:05 <zodbot> Meeting started Thu Sep 24 15:00:05 2009 UTC.  The chair is skalnik1. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:05 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:18 <skalnik1> Hello and welcome everybody
15:01:01 <skalnik1> we should start with recap of last week
15:01:10 <skalnik1> #topic Last week recap
15:02:08 <skalnik1> I introduced tests done by jvcelak - scomes in virtual os 32 and 64 bit
15:02:31 <skalnik1> jvcelak will show more info today
15:03:05 <skalnik1> we discussed test results and we continue in test day preparation
15:04:55 <skalnik1> #topic scomes in virtual os
15:05:31 <jvcelak> Hello everyone :-) My results are on fedora wiki: https://fedoraproject.org/wiki/SIGs/PowerManagementBaseOS/Testcase_Virtualized_32bit_vs_64bit
15:07:27 <jvcelak> Everything is described there. We measured syscall count, tick count, read/written bytes when building squid RPM package.
15:07:52 <jvcelak> The test run 10 times in each environment.
15:09:18 <jvcelak> VirtualBox OSE 32b, 64b. And according to remarks from last meeting I rerun the test in KVM (both 32 and 64b).
15:10:39 <skalnik1> there are nice charts in the report and the first thing to see - stable results on 64 bit systems. KVM seems to be a bit more precise
15:11:33 <jvcelak> Yes. If you are interested and don't like figues you can give a look at the graphs. ;)
15:13:13 <jvcelak> That's probably all I wanted to say to this. If you have some notes, ideas or want me to clarify something about the test - just tell me.
15:13:38 <skalnik1> nice, thanks
15:13:55 <skalnik1> any questions/remarks to this topic?
15:14:59 <jscotka> skalnik1: I'm hier :-)
15:15:20 <skalnik1> hi, i've noticed your entering the room
15:15:22 <skalnik1> :-D
15:15:25 <skalnik1> #topic Test Day Oct-22
15:15:53 <jscotka> skalnik1: one beer for me :-)
15:17:10 <skalnik1> There is available this link https://fedoraproject.org/wiki/Test_Day:2009-10-22
15:17:47 <skalnik1> it's still under construction - we'll update it next days
15:18:17 <skalnik1> there is added tuned-adm test case
15:18:49 <skalnik1> scomes test case is on internal review now - tomorrow should be added to link above
15:20:06 <skalnik1> live cd - hot topic
15:20:18 <jscotka> skalnik1: why hot?
15:20:31 <plautrba> because you have to burn
15:21:44 <jscotka> why not? as a last chance for isntalling
15:23:21 <skalnik1> jscotka: yes, the last chance for installing but common live cd - not with pm "specials"
15:25:55 <skalnik1> we talked internally to make rpm package that will install test cases + dependency
15:27:49 <skalnik1> jscotka: any comments? ;-)
15:28:18 <jscotka> skalnik1: nope :-) only one, on lide CD is sure that dependency are okay :-)
15:30:47 <jvcelak> Are we talking about the live CD as a last chance for installing or running the tests from live-cd?
15:31:01 <skalnik1> yes
15:31:16 <jvcelak> Installing. Perfect. :-) Thanks. I wasn't sure.
15:32:34 <skalnik1> and what about the last chance? or the chance after the last one? :-)
15:33:56 <jscotka> skalnik1: for example? ;-) install freeDOS and recompilation of programs there? :-)
15:36:15 <skalnik1> jvcelak: some progress with live cd?
15:37:44 <jvcelak> Actually... no. :-) My kickstart still fails. I think that this should by done by people from QA. Don't they have experience with it?
15:38:45 <skalnik1> ok, than i will ask dpravec
15:39:02 <skalnik1> #topic documentation and tests
15:42:54 <skalnik1> marcela wrote to Don Domingo about presentation on Development Conference
15:43:39 <skalnik1> she works on document that should be input for Don. This contains good tips for programming in Linux
15:44:36 <skalnik1> we have available notebook with installed RHEL5, F11 and F12
15:44:57 <skalnik1> to compare tests between these systems
15:45:20 <skalnik1> I tried to build bltk for RHEL 5
15:45:20 <jvcelak> skalnik1: Do you have some results now?
15:45:24 <skalnik1> yes
15:45:58 <skalnik1> first result was done - bltk idle test
15:46:08 <skalnik1> RHEL5 x F12
15:46:29 <skalnik1> 10 min test
15:47:14 <skalnik1> RHEL5 consumed 5.08 mWh of energy and F12 4.3 mWh
15:48:15 <skalnik1> this is on an average RHEL5 18.3 W against 15.5 W in the case of F12
15:49:49 <jscotka> skalnik1: niceeee
15:50:15 <skalnik1> we need more results but the difference is quite significant so i decided to make it public
15:50:49 <skalnik1> i have to fix open office workload because it doesn't work on RHEL5
15:51:35 <jscotka> isnt there same problem as I have with other shortcuts?
15:52:39 <skalnik1> there is probably problem with the utilities used by bltk
15:53:02 <jscotka> ah
15:55:22 <skalnik1> well, i have no topic for today
15:58:36 <skalnik1> so thanks everyone for joining and read&write you next week .... ;-)
15:58:44 <skalnik1> #endmeeting