--- Log opened Sun Jul 11 00:00:05 2010
10:50 -!- arete74 [~arete74@net-93-144-185-74.cust.dsl.vodafone.it] has joined #t2
10:55 -!- arete74 [~arete74@net-93-144-185-74.cust.dsl.vodafone.it] has quit [Client Quit]
12:21 -!- mpp [~mpp@2001:470:9a41:107:850a:c37:44f3:f9a2] has joined #t2
12:45 < CIA-29> rene * r36745 /trunk/package/network/cherokee/cherokee.desc: * updated cherokee (1.0.0 -> 1.0.5)
13:06 -!- mpp [~mpp@2001:470:9a41:107:850a:c37:44f3:f9a2] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100625231939]]
17:24 < rxr> re
17:25 < rxr> 37°C olè!
17:26 -!- mjungwirth [~mjungwirt@213.174.252.112] has joined #t2
18:00 -!- mjungwirth [~mjungwirt@213.174.252.112] has quit [Quit: Nettalk6 - www.ntalk.de]
18:32 < rxr> I think with the aircon running the temperature inside our local office machines decreases by about 10° :-)
18:32 -!- mpp [~mpp@2001:470:9a41:107:850a:c37:44f3:f9a2] has joined #t2
18:32 < rxr> hi mpp 
18:32 [Users #t2]
18:32 [@ChanServ] [ koan] [ mjungwirth2] [ mtr] [ Stealth] [ TobiX] 
18:32 [ CIA-29  ] [ LMJ ] [ mpp        ] [ rxr] [ T-One  ] [ zod  ] 
18:32 -!- Irssi: #t2: Total of 12 nicks [1 ops, 0 halfops, 0 voices, 11 normal]
18:32 < mpp> moinsen
18:32 < mpp> somebody turn of the oven outside - please !
18:32 < mpp> LOL
18:33 < rxr> 18:32 < rxr> I think with the aircon running the temperature inside our local office machines decreases by about 10° :-)
18:33 < rxr> I personally blame the global warming
18:34 < mpp> best place to hide these days - at the poles or at server housing facilities
18:34 < rxr> yep, or deep celars
18:34 < mpp> i'm glad you fixed the linux24.conf
18:35 < mpp> thanx for that
18:35 < rxr> last year or so we got this aircon for the office
18:35 < rxr> mpp: welcome
18:35 < mpp> coool
18:35 < rxr> so far we usually could run that aircon in eco mode
18:35 < rxr> it is not a split unit, we can not built that into our rented office rooms, ...
18:35 < rxr> but one of the better non-split units ...
18:36 < rxr> it can use water you fill into it to improve efficience
18:36 < mpp> as long as it does it's job ...
18:36 < rxr> so far the least electric engery consuming ECO mode was sufficient
18:36 < rxr> but, these days even the turbo mode does not really cool down the office that much
18:36 < mpp> any plans so far to release an up to date minimal iso image
18:36 < mpp> mabye -rc3 or something
18:37 < rxr> at leasst the HD and CPU temperatures decrease some 10° :-)
18:37 < mpp> :-)
18:37 < rxr> well, I would hve tagged final long ago if not always some silly tiny crap pilled up again
18:37 < rxr> if my latest cross depmod stuff fixed USB and HID module auto-loading on the images I will probably release 8.0-final
18:38 < rxr> way too long overdue
18:38 < rxr> but then again -EBUSY life
18:38 < mpp> roger that !
18:38 < mpp> unfortunately i missed the big server this week 16core 256gig ram 
18:39 < mpp> it's already deployd
18:39 < mpp> :-(
18:42 < rxr> http://twitter.com/renebln
18:43 < CIA-29> rene * r36746 /trunk/package/develop/ppl/ppl.conf: * helped ppl with auto* missing confusion
18:48 < rxr> load average: 6.15, 7.02, 6.04
18:48 < rxr> t2 loves many-cores to build
18:48 < mpp> yep
18:48 < rxr> need some better networked storage system, though
18:49 < mpp> ?
18:49 < mpp> you mean san / nas ?
18:49 < rxr> some distributed storage
18:49 < mpp> ah
18:49 < rxr> high performance and massive of it
18:49 < mpp> np
18:50 < rxr> what do you recommend ?
18:50 < mpp> in terms of hw or sw ?
18:51 < rxr> you said "np", I'm curious what you can recommend
18:51 < mpp> ah ....
18:51 < mpp> it solely depends on the $$$
18:51 < mpp> how much storage
18:51 < mpp> are we talking about
18:52 < rxr> yeah, I got no fortune to spent on it
18:52 < mpp> i see
18:52 < rxr> I may experiment with some loosely connected google a-like cluster
18:52 < mpp> well simple no bugs no storage
18:52 < mpp> .
18:52 < mpp> :-)
18:53 < mpp> anyways
18:53 < rxr> maybe stuffing some 2-5 TB discs into each VM node and reduandtly spread it over all nodes for a start
18:53 < mpp> get comodity hw
18:53 < mpp> a gigabit switch
18:53 < rxr> yeah, I did :-)
18:53 < mpp> lot's of disks
18:54 < mpp> with nbd you can stripe volumes
18:54 < mpp> or go with openfiler and zfs
18:54 < rxr> zfs? .oO
18:54 < rxr> btw. got these now:
18:54 < rxr> model name	: AMD Phenom(tm) II X6 1055T Processor
18:54 < rxr> wish'ed AMD would deliver the 95W version to me :-)
18:54 < rxr> apparently they only ship it in Japan :-(
18:54 < mpp> that shoud do the job
18:55 < mpp> you need HA for the storage
18:55 < mpp> ?
18:55 < rxr> yep, quite an performance bump over my v1 nodes :-)
18:55 < rxr> I at least need redundant backup to off-site locations
18:55 < mpp> then drbd is the choice
18:56 < rxr> mpp: btw. your famour openssl performance test: http://rene.rebe.de/2010-06-08/x86-64bit-performance-increase/
18:56 < mpp> choose iscsi or nfs for presentation
18:56 < rxr> my v2 nodes have about twice the performance as my v1 hardware setup
18:56 < rxr> but it is about 1 year in between both choices
18:56 < mpp> makes sense
18:57 < mpp> but don't be fooled by openssl benching
18:57 < mpp> real world applications to NOT get boosted automagically with 64bit 
18:57 < rxr> I know
18:57 < rxr> been telling that for a long time
18:57 < mpp> but you knew already i guess
18:57 < rxr> mostly inner loops, and massive data centric applications
18:58 < rxr> but then again thist increas was stunishing
18:58 < mpp> gcc does perform better on t2 with 64bit
18:58 < mpp> BUT not the huge boost one would wish for
19:00 < mpp> okay im off for today
19:00 < mpp> see ya around
19:00 -!- mpp [~mpp@2001:470:9a41:107:850a:c37:44f3:f9a2] has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.6/20100625231939]]
19:00 < rxr> on x86-64 the more than a handful register help a great deal
19:56 -!- danielsh_ [~danielsh@unaffiliated/danielsh] has joined #t2
19:56 -!- danielsh_ [~danielsh@unaffiliated/danielsh] has left #t2 []
19:59 -!- T-One [~T-One@194.183.141.238] has quit [Ping timeout: 245 seconds]
20:00 -!- T-One [~T-One@194.183.141.238] has joined #t2
20:10 < rxr> ach meine herrn
20:10 < rxr> libtool suckt doch
20:10 < rxr> this is soooooooooooooo annoying:
20:10 < rxr> make[1]: Entering directory `/usr/src/t2-trunk/src.exiv2.default.20100711.200858.21585.exactcode.de/exiv2-0.19/xmpsdk/src'
20:10 < rxr> libtool: compile: unable to infer tagged configuration
20:10 < rxr> libtool: compile: specify a tag with `--tag'
20:10 < rxr> make[1]: *** [XML_Node.o] Error 1
20:10 < rxr> make[1]: *** Waiting for unfinished jobs....
20:16 < rxr> pff, manuall adding --tag CC and --tag CXX helps, of course, who the heck invents such useless crap
20:16 < rxr> people should have better stuff to do then writing such br0ken crap, ...
23:57 -!- T-One [~T-One@194.183.141.238] has quit [Ping timeout: 248 seconds]
--- Log closed Mon Jul 12 00:00:05 2010