1335148127 J * fisted_ ~fisted@xdsl-84-44-145-226.netcologne.de
1335148527 Q * fisted Ping timeout: 480 seconds
1335154809 Q * ensc Ping timeout: 480 seconds
1335155360 Q * Alex[fob] Remote host closed the connection
1335156089 M * Bertl off to bed now ... have a good one everyone!
1335156093 N * Bertl Bertl_zZ
1335159319 Q * Radiance Ping timeout: 480 seconds
1335161691 Q * jeroen__ Quit: Ex-Chat
1335161984 J * jeroen__ ~jeroen@095-097-051-172.static.chello.nl
1335162335 J * ncopa_ ~ncopa@3.203.202.84.customer.cdi.no
1335162341 Q * ncopa_ 
1335162364 J * vspas ~vspas@87.213.36.165
1335162639 M * ncopa morning
1335162690 M * ncopa if /etc/vservers/myvserver got deleted, whats the best method to recreate it?
1335162747 M * ncopa i could mv /vservers/myvserver /vservers/my2 and use rsync method, but it feels wrong to conle the data
1335162749 M * ncopa clone*
1335163211 M * daniel_hozac -m skeleton, rm -fr /vservers/myvserver; mv /vservers/{my2,myvserver}
1335163249 M * ncopa i am currently creating the /etc/vservers/myvservers by hand
1335163261 M * ncopa its only 3 symlinks thats needs to be recreated
1335163290 M * ncopa i mean, i am cloning antoher similar vserver config
1335163579 J * ghislain ~AQUEOS@adsl1.aqueos.com
1335167221 M * ghislain hello there
1335167236 M * ghislain oh i see 3.2 and 3.0 new kernels..
1335167317 M * ghislain Bertl_zZ: i do not know if you had time to check the oops and see if it is vserver related or mainline
1335168349 Q * fisted_ Ping timeout: 480 seconds
1335168556 M * ghislain i have the D process lockup in 3.0.26 as well
1335169457 M * ghislain hum false alert on 3.0.26
1335169783 Q * nou Ping timeout: 480 seconds
1335169828 J * nou Chaton@causse.larzac.fr.eu.org
1335172152 M * daniel_hozac Bertl_zZ: i think i know why the bias isn't working
1335174727 M * daniel_hozac real_start_time is boot based, bias_clock is not.
1335174901 J * ghislain1 ~AQUEOS@adsl1.aqueos.com
1335175281 Q * ghislain Ping timeout: 480 seconds
1335175670 J * derjohn_mob ~aj@87.253.171.208
1335178443 Q * Mr_Smoke Read error: No route to host
1335178446 J * Mr_Smoke ~smokey@layla.lecoyote.org
1335179130 J * petzsch ~markus@dslb-088-075-172-224.pools.arcor-ip.net
1335179295 Q * DLange Quit: cul8er
1335179870 J * DLange ~DLange@dlange.user.oftc.net
1335180006 Q * ensc|w Remote host closed the connection
1335180016 J * ensc|w ~ensc@www.sigma-chemnitz.de
1335180317 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com
1335182321 Q * ex Ping timeout: 480 seconds
1335182846 J * ex ex@valis.net.pl
1335183597 Q * urbee Read error: Connection reset by peer
1335183611 J * urbee ~urbee@93-103-199-233.dynamic.t-2.net
1335185611 N * Bertl_zZ Bertl
1335185617 M * Bertl morning folks!
1335189038 Q * BenG Remote host closed the connection
1335189286 J * fisted ~fisted@xdsl-84-44-145-226.netcologne.de
1335189293 Q * fisted 
1335189300 J * fisted ~fisted@xdsl-84-44-145-226.netcologne.de
1335190011 M * ghislain1 morning bertl ^^
1335190822 M * Bertl daniel_hozac: so either we convert to boot based time or drop the VIRT_UPTIME in array.c completely (maybe replace with a sanity check)
1335191211 M * daniel_hozac the only case where it is needed is for the blend through init, right?
1335191237 M * daniel_hozac or do we need to do some offseting to get the dates correct?
1335191253 M * daniel_hozac (i'm not sure how ps calculates it's timestamps)
1335191352 M * Bertl I'm not sure the uptime even affects ps in this case
1335191377 M * Bertl i.e. probably a blend through init would just get a date _before_ the guest system was started
1335191382 M * daniel_hozac right
1335191426 M * Bertl so, we might simply use the start time as lower boundary
1335191465 M * daniel_hozac i'll run some tests when i get home to see what works.
1335191504 M * Bertl excellent! thanks!
1335191720 J * fisted_ ~fisted@xdsl-78-35-88-76.netcologne.de
1335192133 Q * fisted Ping timeout: 480 seconds
1335193113 J * dowdle ~dowdle@scott.coe.montana.edu
1335193587 M * Bertl off for a nap ... kind of tired today ... bbl
1335193603 N * Bertl Bertl_zZ
1335194033 Q * vspas Ping timeout: 480 seconds
1335194977 Q * petzsch Quit: Leaving.
1335195618 J * thierryp ~thierry@zankai.inria.fr
1335195924 Q * derjohn_mob Ping timeout: 480 seconds
1335196568 J * derjohn_mob ~aj@87.253.171.216
1335197209 J * bonbons ~bonbons@2001:960:7ab:0:4d4a:4f19:1fcd:510c
1335197633 Q * derjohn_mob Ping timeout: 480 seconds
1335198875 J * petzsch ~markus@dslb-088-075-172-224.pools.arcor-ip.net
1335200014 Q * thierryp Remote host closed the connection
1335200456 J * imcsk8 ~ichavero@148.229.1.11
1335201346 J * Alex[fob] ~alex@2001:638:80a:109::152:0
1335204097 J * vspas ~vspas@82-169-243-226.ip.telfort.nl
1335206534 Q * vspas Ping timeout: 480 seconds
1335207208 Q * hparker Quit: I've fallen off the 'net and can't get up
1335207275 J * hparker ~hparker@2001:470:1f0f:32c:222:19ff:fe22:b7b5
1335208886 Q * ntrs Quit: Lost terminal
1335209128 J * ntrs ~ntrs@vault08.rosehosting.com
1335209188 J * ViruSzZ ~dejan@mccain.rosehosting.com
1335209232 M * ViruSzZ Greetings there gurus. 
1335209266 M * ViruSzZ Is there anything specific I need to implement in order to run openSUSE 12 as a guest ?
1335209279 M * ViruSzZ I'm getting the following when I'm trying to use yast2
1335209296 M * ViruSzZ The /sys filesystem is not mounted. If you are running in a chroot environment, bind-mount missing filesystems.
1335209314 M * daniel_hozac lol
1335209331 M * daniel_hozac yast requires /sys?
1335209364 M * ViruSzZ yes it seems it requires.
1335209370 J * derjohn_mob ~aj@c140030.adsl.hansenet.de
1335209407 N * Bertl_zZ Bertl
1335209410 M * Bertl back now ...
1335209439 Q * ghislain1 Quit: Leaving.
1335209443 M * Bertl ViruSzZ: what does it use /sys for?
1335209543 M * daniel_hozac with yast, i doubt anyone really knows.
1335209545 M * ViruSzZ I have no idea why yast2 needs /sys for. Anyhow, I'm trying to execute 
1335209547 M * ViruSzZ  # yast --help
1335209547 M * ViruSzZ The /sys filesystem is not mounted.
1335209547 M * ViruSzZ If you are running in a chroot environment, bind-mount missing filesystems.
1335209605 M * Bertl yeah, well, you can do that, but be aware that sys will drastically reduce your security
1335209630 M * Bertl but I'd use strace to first figure out what it actually wants from there
1335209648 M * Bertl because it might be simpler to put a tmpfs there with a few dummy files
1335209660 M * daniel_hozac or safer for sure.
1335209680 M * ViruSzZ searching the web I found this => http://lists.opensuse.org/yast-commit/2009-04/msg00343.html
1335209682 M * ViruSzZ +test -z `ls /sys |head -n1` && echo "The /sys filesystem is not mounted." &&
1335209682 M * ViruSzZ mount_ok=0;
1335209695 M * ViruSzZ so does this means that if /sys is empty yast will just exit?
1335209716 M * daniel_hozac yes
1335209732 M * daniel_hozac touch /sys/file would probably trick it.
1335209742 M * ViruSzZ :D hahaha what the ... i just created empty file
1335209746 M * ViruSzZ and it seems to work now
1335209944 M * daniel_hozac 3.3.2-vs2.3.3.2 is a lot of threes and twos.
1335210085 M * Bertl we'll have 3.3.3-vs2.3.3.3 soon :)
1335210308 M * daniel_hozac couldn't we name it 3.3.3-vs3.3.3.3?
1335210367 M * Bertl :)
1335210783 M * daniel_hozac hmm, 3.3.2-vs2.3.3.2 fails to compile here, due to missing #include "pnode.h" in fs/namei.c
1335210831 M * Bertl well, it definitely compiled here, something config specific?
1335210876 M * daniel_hozac could be -Werror?
1335210884 M * daniel_hozac but it seems to be a legit warning.
1335210895 M * daniel_hozac mnt_get_count is no longer public, and its definition has changed.
1335210922 M * daniel_hozac (but that is only used in CoW debug messages, so it shouldn't cause any problems...)
1335210989 M * Bertl okay, anyway, we should add the include
1335211019 M * daniel_hozac that causes multiple includes and redefinitions here...
1335211083 M * Bertl alternatively we can move the debug into a different place or make mnt_get_count public again
1335211111 M * Bertl (or drop the debug output completely, but I guess that's the worst solution)
1335211309 J * Educated ~dejan@77.28.99.153
1335211637 Q * FireEgl Quit: Leaving...
1335211773 Q * ViruSzZ Ping timeout: 480 seconds
1335211794 Q * Educated Ping timeout: 480 seconds
1335212603 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-cow-fix28.diff builds for me.
1335213672 J * FireEgl FireEgl@2001:470:e056:1:495e:ef1:cf34:2162
1335214026 Q * bonbons Quit: Leaving
1335214898 J * geb ~geb@mars.gebura.eu.org
1335215389 Q * Alex[fob] Ping timeout: 480 seconds
1335215654 Q * ntrs Ping timeout: 480 seconds
1335215784 Q * fisted_ Ping timeout: 480 seconds
1335216019 J * fisted ~fisted@xdsl-81-173-189-226.netcologne.de
1335216166 J * ntrs ~ntrs@vault08.rosehosting.com
1335217328 Q * petzsch Quit: Leaving.
1335218003 M * daniel_hozac Bertl: hmm, i can't make any sense of this.
1335218016 M * daniel_hozac a vxdprintk(VXD_CBIT(misc, 4), "start_time = %llu, bias = %llu", start_time, bias); in the virt_uptime block gives me
1335218029 M * daniel_hozac [  507.554590] vxD: ffff880037f21740: start_time = 36524, bias = 25207
1335218049 M * daniel_hozac yet, the start_time field is 0
1335218086 M * daniel_hozac (i.e. awk '{ print $21 }' /proc/<pid>/stat shows 0)
1335218229 M * daniel_hozac ... because $22 is the correct field :-)
1335218340 M * Bertl :)
1335218378 M * daniel_hozac counting is hard sometimes.
1335218684 M * Bertl off by one errors are a quite common problem ...
1335219038 M * daniel_hozac hmm. i've tried following ps's code, and doing it manually. however, that seems to be getting the correct result...
1335219114 M * daniel_hozac http://paste.linux-vserver.org/20953
1335219126 M * daniel_hozac all executed in rapid succession (i.e. less than a second apart)
1335219250 M * Bertl so what we know so far is that 'certain' ps gives the wrong results ...
1335219263 M * Bertl s/gives/give/
1335219344 M * daniel_hozac (the delta between real time and ps perceived time is the same as the bias)
1335219672 M * daniel_hozac hmm
1335219761 M * daniel_hozac it seems newer procps use btime from /proc/stat as the boot time
1335219776 M * daniel_hozac instead of calculating it based on uptime
1335219803 M * daniel_hozac so we're missing a VIRT_UPTIME check in fs/proc/stat.c:show_stat
1335219835 M * Bertl ah, yes, I see
1335219923 M * Bertl maybe we should modify getboottime
1335220076 M * daniel_hozac perhaps. is it used anywhere else?
1335220101 M * Bertl several places, but nothing really relevant to a guest
1335220283 M * daniel_hozac might be safer then to just add the check there?
1335220290 M * daniel_hozac in show_stat, i mean.
1335220305 M * Bertl yeah, probably
1335220350 M * Bertl if (vx_flags(VXF_VIRT_UPTIME, 0))
1335220368 M * Bertl   vx_vsi_uptime(&boottime, NULL);
1335220464 M * daniel_hozac wouldn't that adjust it the wrong way?
1335220540 M * daniel_hozac isn't jif += current->vx_info->cvirt.bias_uptime.tv_sec; what we'd like?
1335220594 M * Bertl not sure, you tell me, you did the tests :)
1335220617 M * daniel_hozac i'm testing mine now, so we'll see :)
1335220654 M * daniel_hozac looks like it's correct.
1335220728 M * daniel_hozac http://paste.linux-vserver.org/20954
1335220734 M * daniel_hozac guest started about two minutes after the host
1335220805 M * Bertl okay, so we need correction in both directions
1335220837 M * daniel_hozac yeah
1335220841 M * Bertl let's introduce a vx_vsi_boottime then
1335220851 M * daniel_hozac sounds good
1335220853 M * Bertl I'll prepare a patch
1335220868 M * Bertl (unless you like to do it of course :)
1335220873 M * daniel_hozac i'm gonna head to bed, but i'll give it a spin when i get up.
1335220885 M * Bertl okay, great, thanks!
1335220901 M * daniel_hozac have a good one
1335221435 M * Bertl you too, sleep well
1335221620 Q * dowdle 
1335222893 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1335224621 J * DoberMann_ ~james@2a01:e35:8b44:84c0::2
1335224622 Q * DoberMann Remote host closed the connection
1335224993 Q * DoberMann_ Remote host closed the connection
1335224995 J * DoberMann ~james@2a01:e35:8b44:84c0::2