1329352891 Q * dowdle 
1329353150 M * Bertl off for a nap now ... bbl
1329353162 N * Bertl Bertl_zZ
1329357664 J * ghislain ~AQUEOS@adsl2.aqueos.com
1329358843 Q * ghislain Read error: Connection reset by peer
1329361002 N * Bertl_zZ Bertl
1329361010 M * Bertl back now ...
1329372267 Q * imcsk8 Ping timeout: 480 seconds
1329372353 Q * bergerx Quit: Leaving
1329373227 Q * cehteh Read error: Connection timed out
1329377405 J * imcsk8 ~ichavero@148.229.1.11
1329378474 J * gucki ~gucki@80-218-125-247.dclient.hispeed.ch
1329379432 J * ghislain ~AQUEOS@adsl2.aqueos.com
1329381340 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com
1329381763 N * BobR_zZ BobR
1329381897 J * xe ex@valis.net.pl
1329381897 Q * ex Read error: Connection reset by peer
1329381907 N * xe ex
1329382602 Q * ensc|w Remote host closed the connection
1329382611 J * ensc|w ~ensc@www.sigma-chemnitz.de
1329382767 J * ensc ~irc-ensc@p54ADE36B.dip.t-dialin.net
1329383611 N * BobR BobR_afk
1329383809 J * uranus ~uranus@pD9E5AB54.dip.t-dialin.net
1329383817 M * uranus hi all
1329383833 M * uranus Bertl, what was changed between vs2.3.2.2 and vs2.3.2.3?
1329383956 M * Bertl a config check, nothing really important
1329383966 M * uranus ok
1329384455 M * ghislain woot hello there
1329384477 M * ghislain i just done a partprobe on a server and got a  kernel:Oops: 0010 [#1] SMP
1329384486 M * ghislain kernel:last sysfs file: /sys/block/vroot0/queue/physical_block_size
1329384530 M * ghislain it's on an old kernel but just to notice that:  2.6.33.2-vs2.3.0.36.30.4
1329384809 M * Bertl could you upload the full output?
1329385304 N * BobR_afk BobR
1329385543 M * ghislain yes: http://paste.linux-vserver.org/20736
1329385551 M * ghislain but no debug kernel enable :p
1329385611 M * ghislain basicaly i was doing a change on /boot, reduce  it so i have enough gap between start of disk and part1 to have grub2 installed for squeeze upgrade
1329385632 M * ghislain when i done a partprobe to make the kernel notice the change it crashed
1329385659 M * ghislain it oopsed to be more precise i could still log in ssh
1329385722 M * ghislain i stopped the vserver guest before so in theory vroot should have been disabled ( "should"(tm) )
1329385804 M * ghislain i done the same on some recent kernel without issue but as a tree could hide a forest i just told you "in cas"
1329385807 M * ghislain case
1329385841 M * Bertl well, looks strange, sure that is the first 'BUG' message you got?
1329385876 M * Bertl it seems that the kernel managed to transfer execution to '0' (as address) which is really strange
1329385926 M * Bertl i.e. looks like a low level kernel bug and/or a hardware issue to me
1329385934 M * ghislain yes this is the first as far as the dmesg  show i have only   did lookup hidden proc like usual and then the /dev/md0 stop and then the BUG
1329385966 M * ghislain i have some : vcontext[15785]: segfault at 7fff4295bff8 ip 0000000000402f23 sp 00007fff4295c000 error 6 in vcontext[400000+8000]
1329385972 M * ghislain before that toot
1329385984 M * Bertl that shouldn't happen either IMHO
1329386044 M * ghislain ok for now i try to bring it back, as i must do this on a bunch of servers i will tell you if i encounter it again and if i can have more info
1329386193 M * Bertl please do so, also make a note about the kernel versions
1329386193 M * ghislain i have 10 segfault in the dmesg, humm i think i will start to monitor more closely this kern.log i should have see those
1329386198 M * ghislain yes
1329386230 M * ghislain do you see a way to debug those segfault on a standard kernel ( i mean with no debug loaded)
1329386234 M * Bertl the vcontext segfaults do not need to be related, they could be caused by a dietlibc bug
1329386280 M * Bertl well, you can lookup the offsets given in the call trace
1329386308 M * Bertl e.g. find the base of generic_make_request, add 0x193 to it, check which line that corresponds to
1329386349 M * Bertl but unless you see that on more than one machine, I'd blame the hardware
1329386375 M * ghislain i dont know how to do that, if you explain i can write a wiki faq so you can relate people to it for bug report. hum i havent looked if there is not allready one !
1329386401 M * ghislain i am doing a search in the logs of all my server right now to see if i have other case
1329386426 M * daniel_hozac could be that vroot is missing some hook that partprobe uses?
1329386466 M * Bertl could be, that would actually be a quite good explanation for a jump to 0
1329386494 M * daniel_hozac it does seem to be trying to flush it.
1329386795 M * Bertl blkdev_issue_flush checks for a make_request_fn
1329386839 M * Bertl what was the reason we allowed the make_request on vroot?
1329386959 M * ghislain on 30+ servers i have only 2 with vcontext segfault this one and another one with only 1 segfautl since ages
1329386961 M * daniel_hozac IIRC something required it...
1329386999 M * Bertl hmm, well, in theory, we should handle the request properly
1329387043 M * daniel_hozac not on 2.6.33 though :)
1329387082 M * Bertl I don't have a 2.6.33 tree anymore :)
1329387084 M * ghislain on 3.x i have done similar thing without crashing but i cannot tell if i need a special condition or not to trigger it
1329387101 M * ghislain that was on this particular one and not on the 3.x
1329387138 M * ghislain bertl: should be near the commodre64box in the garage ;p
1329387155 M * Bertl yeah, right :)
1329387181 M * daniel_hozac i have back to 2.6.16 on this computer...
1329387191 M * Bertl well, I'd say, daniel_hozac explanation (as usual) is right on spot ... 
1329387197 M * Bertl +'s
1329387230 M * Bertl if you feel like verifying that, you could boot that specific kernel again and run the flush on a configured vroot device
1329387278 M * ghislain a generic partprobe could do that or do i need a more specific command ?
1329387298 M * ghislain i can try it it just must be up at 20h (it is a backup server )
1329387312 M * ghislain so i have 8h to play with it
1329387342 M * daniel_hozac patrprobe should be sufficient
1329387692 M * ghislain ok i have to recover it then i will try it
1329387723 N * ensc Guest2791
1329387733 J * ensc ~irc-ensc@p54ADDE22.dip.t-dialin.net
1329388137 Q * Guest2791 Ping timeout: 480 seconds
1329389939 N * BobR BobR_afk
1329391486 Q * BenG Quit: I Leave
1329391652 M * ghislain the server is doing a fsck after it finishes i will be able to test
1329391714 M * ghislain ok done partprobe crashed again
1329391741 M * ghislain http://paste.linux-vserver.org/20737 same thing :p
1329391772 M * ghislain i am still logged i go eat but if you want me to run some debug please leave a msg i will do it when i come back
1329391895 M * ghislain update: http://paste.linux-vserver.org/20739
1329392463 M * Bertl daniel_hozac: the last paste reminds me of something :)
1329392535 Q * Aiken Remote host closed the connection
1329393900 M * daniel_hozac yeah, that looks about right...
1329395183 N * BobR_afk BobR_zZ
1329395245 M * ghislain bertl: what does it reminds you off, the previous one ? :)
1329396047 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com
1329396122 M * Bertl the 'strange and unexplained' kernel crash where no stack trace was shown (and the fact that nobody could recreate it)
1329397065 M * ghislain this sound so mysterious, is it the title of a book or film ?
1329397812 M * ard I've reviewed that movie over and over...
1329397817 M * ard and I still do not like it :-)
1329397859 M * ard Actually, my unexplained crashes shows no trace, no whatsoever, just a dead machine :-)
1329397867 M * ard but that was so 2011 :-)
1329397944 M * Bertl ard: what kernel version was that?
1329398307 M * ard I think around 2.6.33... 2.6.36
1329398323 M * ard From 2.6.37 I usually got fs wait problems
1329398330 M * ard from 3.1.4 I have stability :-)
1329398370 M * ard none really vserver related, but thanks to vserver we really got to pound the machine :-)
1329398395 M * ard mysql in one vserver doing I/O and lots of memory, php on the other, doing lots of network and cpu
1329398414 M * ard In 2.6.37 it will usually turns out into a divide by zero :-)
1329398445 M * Bertl so that would match the 2.6.33 we have here .. do you use any vroot devices and programs (like the partbrobe) which flush them?
1329398456 M * ard nope...
1329398486 M * ard vserver with ncontext in a shared network namespace with a vlan holder
1329398507 M * ard multiple of those (multiple network namespaces) on a single host
1329398580 M * ard But from what I know is that there seems to be network buffer problems, especially with e1000 and jumbo frames (we use bnx2 and no jumbo frames actuall ;-) )
1329398645 M * ard the network seems to allocate huge amounts of memory in interrupt time, which eats all min-free-memory and then the disk I/O has no memory to do stuff
1329398698 M * ard We do have a firewall running on 2.6.34 which occasionally has a hickup...
1329398741 M * Bertl 2.6.34 mainline or 2.6.34 with Linux-VServer?
1329398755 M * ard mainline, no vserver
1329398785 M * Bertl okay, so despite the similarity, not really related ... well it was just a guess ...
1329398790 M * ard Heh, I see due to the reboots, I even installed a newer kernel which has vserver support
1329398971 M * ard Hmmmm, at least ghislain has an oops ;-)
1329399098 M * ard The question is, is it really vserver related. Will it do the same with vanilla 2.6.33.2
1329399815 M * ghislain ard: while i have some kernel panics sometime in this case i have the oops but the machine still run 
1329400372 M * ghislain but i also encoutnter kernel panic  from time to time on some servers
1329400402 M * ghislain but they happen at some month interval and i do not have the trace or anything as the hoster just reboots the pc
1329400681 M * ard ah :-)
1329400902 Q * michal Ping timeout: 480 seconds
1329401470 M * Bertl off for a nap ... bbl
1329401475 N * Bertl Bertl_zZ
1329401531 M * ghislain after your nap if you want me to run some debug command on this host i got 4h left before needing to make it up again
1329401552 M * Bertl_zZ na, no need, I consider this one resolved
1329401654 M * ghislain ok
1329401656 J * michal ~michal@91.228.56.242
1329401861 Q * grobie Ping timeout: 480 seconds
1329402528 J * Mr_Smoke ~smokey@layla.lecoyote.org
1329402545 M * Mr_Smoke Darn, just missed Bertl_zZ 
1329402548 M * Mr_Smoke morning :)
1329402561 M * Mr_Smoke I'm having trouble (re)-compiling util-vserver with gcc-4.5
1329402599 M * Mr_Smoke I get this: undefined reference to `__stack_chk_fail_local'
1329402605 M * Mr_Smoke I'm running hardened gentoo
1329402613 M * Mr_Smoke (gcc is hardened too)
1329402656 M * Mr_Smoke More precisely /usr/lib/gcc/i686-pc-linux-gnu/4.5.3/../../../../i686-pc-linux-gnu/bin/ld: src/check-unixfile: hidden symbol `__stack_chk_fail_local' isn't defined
1329402664 M * Mr_Smoke /usr/lib/gcc/i686-pc-linux-gnu/4.5.3/../../../../i686-pc-linux-gnu/bin/ld: final link failed: Bad value
1329402710 M * daniel_hozac do you have a recent dietlibc?
1329402769 M * Mr_Smoke Oh hangon, I had pasted something similar about gcc 4.4 about a year ago, haha
1329402772 M * Mr_Smoke google is messing with me
1329402774 M * Mr_Smoke hi daniel_hozac 
1329402776 M * Mr_Smoke Lemme check
1329402783 M * Mr_Smoke ahem
1329402785 M * Mr_Smoke 0.30
1329402789 M * Mr_Smoke let's update
1329402789 M * Mr_Smoke :)
1329402832 M * Mr_Smoke And yeah, you just provided me with the same answer Herbert did that time :)
1329402836 M * Mr_Smoke Sorry about the noise ;)
1329402847 M * Mr_Smoke daniel_hozac: have you heard of/investigated the issue with openrc too ?
1329403044 M * daniel_hozac i was under the impression we solved it.
1329403067 M * Mr_Smoke oh?
1329403078 M * Mr_Smoke I haven't found information relative to that fix then
1329403363 M * Mr_Smoke at least the latest util-vserver still have a problem with the default openrc settings
1329403368 M * Mr_Smoke (ie nothing regarding rc_cgroups)
1329403530 M * daniel_hozac did you set /etc/vservers/.defaults/cgroup/mnt to /sys/fs/cgroup and touch /etc/vservers/.defaults/cgroup/per-ss?
1329403563 M * Mr_Smoke hm not that I recall
1329403585 M * Mr_Smoke I must have missed that somehow :/
1329403714 Q * BenG Quit: I Leave
1329403908 M * Mr_Smoke daniel_hozac: where is this documented btw ?
1329403967 M * Mr_Smoke Hm, flower page mentions the first part
1329404021 M * Mr_Smoke What's that per-ss setting do ?
1329404076 M * Mr_Smoke Ah, per subsys
1329404331 M * Mr_Smoke Maybe I will suggest that hollow adds dietlibc as a dependency to util-vserver
1329404336 M * Mr_Smoke I mean a particular version
1329404435 M * Mr_Smoke daniel_hozac: I'm getting a load of output from util-vserver, but it still won't start
1329404448 M * Mr_Smoke mount: vserver already mounted or /sys/fs/cgroup/cpuset busy
1329404449 M * Mr_Smoke mount: according to mtab, cpuset is already mounted on /sys/fs/cgroup/cpuset
1329404459 M * Mr_Smoke and so on for various cgroup subsets
1329404514 M * Mr_Smoke What am I missing?
1329404636 M * Mr_Smoke In fact, I'm at the same point as this guy : https://bugs.gentoo.org/show_bug.cgi?id=401203
1329404853 M * daniel_hozac try http://paste.linux-vserver.org/20741
1329404929 M * ghislain bertl_zZ: just one question is that problem present in just a few version or in all < 2.6.33 ?
1329404958 M * Mr_Smoke Lemme see
1329405080 M * ghislain daniel_hozac: is there a way when vroot is nto in module to desactivate it so the oops i encountered could be avoided ?
1329405396 M * Mr_Smoke daniel_hozac: nope, didn't help :/
1329405425 M * Mr_Smoke The result is exactly the same
1329405466 M * Mr_Smoke daniel_hozac: distrgard that
1329405468 M * Mr_Smoke IT WORKS :)
1329405479 M * Mr_Smoke (stupid endline issue when downloading the patch)
1329405624 M * Mr_Smoke Thanks a bunch :)
1329405635 M * Mr_Smoke Hope it will result in a new ebuild for util-vserver 
1329405690 M * Mr_Smoke I've commented the bug mentioned above so that everyone can benefie
1329405692 M * Mr_Smoke benefit
1329405693 M * Mr_Smoke cheers !
1329407174 J * dowdle ~dowdle@scott.coe.montana.edu
1329407257 J * gucki_ ~gucki@80-218-125-247.dclient.hispeed.ch
1329407309 Q * gucki Remote host closed the connection
1329407322 Q * gucki_ Remote host closed the connection
1329407341 J * gucki ~gucki@80-218-125-247.dclient.hispeed.ch
1329410404 Q * ncopa Quit: Leaving
1329411835 J * bonbons ~bonbons@2001:960:7ab:0:fc0e:1f0f:5822:bd78
1329412331 J * barismetin ~barismeti@aegis.CS.Princeton.EDU
1329413747 Q * gucki Remote host closed the connection
1329413904 J * bergerx ~bergerx@95.6.16.132
1329414638 Q * Alex[fob] Ping timeout: 480 seconds
1329414663 J * grobie ~grobie@tyr.schnuckelig.eu
1329416364 M * daniel_hozac ghislain: don't run partprobe :)
1329416369 M * daniel_hozac ghislain: aside from that, no.
1329416552 M * ghislain lol
1329416576 M * ghislain i need it to have grub2 installed i need to resize my 1st partition to let 1Mib free at the start
1329416605 M * ghislain i cannot reboot before installing grub, grub will not work before i change the partition table :p
1329416627 M * ghislain this bug was also on the 2.2 vserver version ?
1329416636 M * daniel_hozac yeah.
1329416661 M * ghislain dam it, do you remember at which version it has been dealt with ?
1329416690 M * ghislain at least in 3.2.6 i have not this issue :p
1329416729 M * daniel_hozac looks like patch-2.6.38-vs2.3.0.37-rc8.diff
1329416782 M * ghislain dam it most of my servers runs older than this one, seems i am doomed :p
1329416809 M * daniel_hozac well, it's possible that it was backported to older kernels too.
1329416832 M * daniel_hozac but the fix is from march 2011.
1329416984 M * ghislain ok, thanks a lot for the information. I do not know how i will get out of this lol
1329417007 M * ghislain if i can unload mod vroot freom kernel it will not do that ?
1329417018 M * daniel_hozac right
1329417040 M * ghislain the issue is that older server had the vroot in kernel not in module but at least for "not so old ones" i can do that
1329417045 M * ghislain thx :)
1329417109 M * daniel_hozac if you have support for using e.g. systemtap you could likely work around it.
1329417136 M * daniel_hozac if it's critical that it does not happen.
1329418023 M * ghislain it is not really completly critical, the issue is that i need a new kernel so i need new tools but the new tools require debian squeeze that require grub2 that require me to have a gap between mbr and 1st partition
1329418065 M * ghislain the debian package included in the tool require packages that are from squeeze 
1329418077 M * ghislain a solution will be to compile my own version
1329418082 M * ghislain of the tool
1329418098 M * ghislain then rebboot on new kernel, then i can use partprobe
1329418123 J * alpha_one_x86 ~kvirc@201.222.115.13
1329418165 M * ghislain i have to add a step to compile an intermediate tool
1329418293 M * alpha_one_x86 Hello, I have into my cgroup folder for my guest (/dev/cgroup/vmX/) just task on some guest, and other all to control a full cgroup, I assume the difference as with/without file into the /etc/vserver/vmX/cgroup/, rights?
1329418374 J * uranus_ ~uranus@pD9E5AE3E.dip.t-dialin.net
1329418382 M * daniel_hozac alpha_one_x86: what?
1329418431 M * alpha_one_x86 If I don't put file into cgroup folder, my vserver have not cgroup process controler enabled?
1329418491 M * alpha_one_x86 it seam not :(
1329418498 M * daniel_hozac depends on your kernel.
1329418551 M * alpha_one_x86 Why on some host, I have well cgroup on my vserver, and on some other just "task" file into the cgroup folder? I have controled the kernel
1329418574 M * daniel_hozac you probably didn't enable any subsystems.
1329418579 M * daniel_hozac you can check with cat /proc/cgroups
1329418821 Q * uranus Ping timeout: 480 seconds
1329418940 M * alpha_one_x86 http://pastebin.com/7y1iQjT6 -> seam correct?
1329419732 M * daniel_hozac what are your mount options?
1329420529 M * alpha_one_x86 none                            /dev/cgroup             cgroup  cpu,cpuset,memory,noauto in all case
1329420559 M * alpha_one_x86 (Server installed from same archive, only guest change I think)
1329421989 Q * bergerx Quit: Leaving
1329423339 Q * imcsk8 Ping timeout: 480 seconds
1329423879 J * hijacker_ ~hijacker@cable-84-43-136-96.mnet.bg
1329423975 N * Bertl_zZ Bertl
1329423980 M * Bertl back now ...
1329425976 Q * hijacker_ Quit: Leaving
1329426429 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1329426849 J * imcsk8 ~ichavero@148.229.1.11
1329427076 Q * cuba33ci Read error: Connection reset by peer
1329427176 J * cuba33ci ~cuba33ci@114-36-232-84.dynamic.hinet.net
1329427341 Q * uranus_ Remote host closed the connection
1329428221 J * sweil ~stefan@p5086F90E.dip.t-dialin.net
1329428783 Q * FireEgl Quit: Leaving...
1329429804 Q * bonbons Quit: Leaving
1329430141 J * petzsch ~markus@p57B67DE6.dip.t-dialin.net
1329431891 M * alpha_one_x86 I thinks it's vserver + cgroup bug
1329432536 Q * barismetin Remote host closed the connection
1329432767 Q * petzsch Quit: Leaving.
1329433025 M * ghislain do you know why the debian packages of the tools require diffutils ?  as it does not exist on lenny i cannot build them for lenny
1329433046 Q * sweil Remote host closed the connection
1329433234 M * Bertl ghislain: are you building from sources?
1329433253 M * Bertl alpha_one_x86: that your cgroup system is mounted incompletely? how so?
1329433369 M * ghislain bertl: yes i use the source and use the included debian build block
1329433394 M * alpha_one_x86 it seam be fully mounted, and all part (cpu, memory ..) is mounted: http://pastebin.com/uWJL80LV
1329433406 M * alpha_one_x86 I will try update the kernel and vserver patch
1329433417 M * Bertl what's your current kernel/patch?
1329433452 M * alpha_one_x86 3.1.10 and vs2.3.2.5
1329433471 M * Bertl that should be more than fine
1329433479 M * Bertl util-vserver version?
1329433497 M * alpha_one_x86 0.30.216_pre2955
1329433507 Q * hparker Quit: I've fallen off the 'net and can't get up
1329433512 M * Bertl that's a little older, but should be fine as well, IIRC
1329433537 M * Bertl and only some guests have no cgroup association?
1329433577 M * alpha_one_x86 no, it's by host, on one host, all guest have it, on other any have it
1329433611 M * Bertl s/any/some/? or s/any/none/?
1329433630 M * alpha_one_x86 on same host config (same vserver + kernel + vserver utils), one have it, other have not
1329433653 M * alpha_one_x86 *none, sorry for my english
1329433662 M * Bertl no problem, just clarifying
1329433677 M * Bertl okay, so you have two hosts, similar/identical setuo
1329433680 M * Bertl *setup
1329433705 M * Bertl and on one host, no guest has the proper cgroups, but the host shows them mounted?
1329433734 M * alpha_one_x86 yes
1329433745 M * alpha_one_x86 I search the difference, but I found nothing
1329433833 M * Bertl could you upload 'ls /dev/cgroup' for both hosts?
1329433944 M * alpha_one_x86 Working: http://pastebin.com/pNDgWmKd not working: http://pastebin.com/nYLC2SUu
1329434029 M * Bertl yeah, so it looks like it wasn't properly mounted
1329434063 M * Bertl could you also provide the output of 'grep cgroup /proc/mounts' for both?
1329434137 M * daniel_hozac ghislain: what provides the diff and cmp programs on lenny then?
1329434198 M * alpha_one_x86 previouly pasted: http://pastebin.com/uWJL80LV
1329434211 M * alpha_one_x86 I repaste it
1329434235 M * Bertl that's on the non working host, yes?
1329434289 M * alpha_one_x86 http://pastebin.com/nPWUrfC5
1329434330 M * Bertl there you go, that's the difference
1329434354 M * Bertl on the non working _something_ mounts the cgroup subsystem to /sys/fs/
1329434377 M * Bertl on the working one, the util-vserver runlevel scripts mount it to the proper place (/dev/cgroup)
1329434427 M * Bertl stop whatever is mounting it before util-vserver from doing so, and the non working host will work after the next reboot
1329434455 M * alpha_one_x86 it very complex to reboot in my case
1329434476 M * daniel_hozac you could also set util-vserver to use /sys/fs/cgroup
1329434479 M * Bertl yeah, well, unmounting the cgroup mounts and running the util-vserver runlevel scripts should work
1329434498 M * Bertl daniel_hozac: does that work for separate mounts as well?
1329434521 M * Bertl i.e. the setup alpha_one_x86 has (on the non working host) seems to use separate mounts for each subsystem
1329434624 M * alpha_one_x86 the /dev/cgroup is with option noauto, drop it can solve my problem?
1329434675 M * Bertl you are probably talking about an fstab entry, yes?
1329434690 M * alpha_one_x86 yes
1329434708 M * Bertl as I said, by default, the runlevel scripts handle that
1329434726 M * Bertl but of course, you could also do the mount yourself
1329434735 M * Bertl (but you asked for the difference between the hosts :)
1329434799 M * alpha_one_x86 I my case, lot of service is customized (but same way on both server)
1329434837 M * Bertl well, obviously there is (or at least was at boot time) a difference in the setup
1329434865 M * Bertl anyway, doesn't look to me like cgroup _or_ Linux-VServer related, just a config issue
1329434885 M * alpha_one_x86 I think, because all services are start as paralle, then if one have more slow hdd/cpu, the order will change
1329434905 M * alpha_one_x86 I will search about that's, thanks for your help
1329434917 M * Bertl could be, that's why proper system init does impose some 'ordering'
1329435090 M * alpha_one_x86 I have found, I one case, openrc 0.8.3-r1 is used (work), in other case 0.9.8.2 is used (not work)
1329435124 M * Bertl there you go 
1329435155 M * alpha_one_x86 Thanks for you help, I will ask into my distro specific channel
1329435224 M * Bertl you're welcome!
1329435312 Q * alpha_one_x86 Quit: KVIrc KVIrc Equilibrium 4.1.3, revision: 6066, sources date: 20120101, built on: 2012-02-13 10:45:12 UTC http://www.kvirc.net/
1329435474 M * ser it seems 3.0.20-vs2.3.2.2-beng is finally stable, first in 3.x series