1197504070 J * cedric ~cedric@rny93-2-82-66-66-30.fbx.proxad.net 1197504174 Q * edeschen Remote host closed the connection 1197504338 Q * cedric 1197504344 J * ViRUS ~mp@p57A6DA97.dip.t-dialin.net 1197504615 Q * fs Ping timeout: 480 seconds 1197504664 J * fs fs@213.178.77.98 1197504875 Q * mnemoc Ping timeout: 480 seconds 1197505030 J * mnemoc ~amery@kilo105.server4you.de 1197505097 Q * atterdag Ping timeout: 480 seconds 1197505170 J * Infinito argos@200-140-154-162.gnace701.dsl.brasiltelecom.net.br 1197507256 J * edeschen ~edeschen@proxy-sjc-2.cisco.com 1197507480 Q * dowdle Remote host closed the connection 1197507525 Q * Abaddon Quit: leaving 1197507795 N * DoberMann[PullA] DoberMann[ZZZzzz] 1197507846 J * amaxik ~a@r5ct9.net.upc.cz 1197507858 M * amaxik hi all ) 1197507887 M * amaxik i have a some q 1197507978 M * amaxik is it ok to have 2 mounts for / ? 1197507980 M * amaxik like 1197507981 M * amaxik rootfs / rootfs rw 0 0 1197507998 M * amaxik and /dev/root / ext3 rw,tag,data=ordered 0 0 1197508013 M * amaxik why 2 mounts for / ? 1197508383 Q * yarihm Quit: Leaving 1197509051 M * edeschen hi daniel 1197509092 M * edeschen I see that "vserver stop" instruct init running in the vserver to get to runlevel 6 (which I though was for reboot, and 0 for shutting down/halt) 1197509149 M * edeschen another interesting thing is from inside the vserver, runlevel command always return "unknown" 1197514660 Q * Infinito Quit: Quitte 1197515403 J * ntrs__ ~ntrs@79.125.241.214 1197515562 Q * mire Ping timeout: 480 seconds 1197515834 Q * ntrs_ Ping timeout: 480 seconds 1197515887 J * infowolfe ~infowolfe@c-71-193-153-166.hsd1.or.comcast.net 1197518512 Q * _gh_ Ping timeout: 480 seconds 1197520486 Q * shuri Quit: This computer has gone to sleep 1197520608 Q * balbir Ping timeout: 480 seconds 1197521942 J * _gh_ ~gerrit@1406chost40.starwoodbroadband.com 1197524891 Q * ViRUS Quit: Leaving 1197525411 J * undefined ~undefined@adsl-68-94-183-129.dsl.rcsntx.swbell.net 1197525500 P * undefined 1197525601 Q * edeschen Remote host closed the connection 1197526001 J * otaku42 ~otaku42@torvalds.h4ckr.net 1197526004 M * otaku42 moin 1197526064 M * otaku42 maybe a dumb question, but anyway: how do i find out which version of linux-vserver i'm currently using on a host? without being able to look at the sources the current kernel has been compiled from, i mean? 1197526146 M * daniel_hozac uname. 1197526234 M * otaku42 daniel_hozac: that (uname -a) just tells me that the kernel has vserver extensions compiled in, but not which version they are: Linux marvin 2.6.22-2-vserver-686 #1 SMP Fri Sep 21 15:20:55 UTC 2007 i686 GNU/Linux 1197526268 M * otaku42 daniel_hozac: the host is on debian etch, and the kernel is from backports.org 1197526281 M * daniel_hozac so... don't use Debian kernels? :) 1197526320 M * daniel_hozac usually the changelog has the version... 1197526411 M * otaku42 daniel_hozac: the changelog suggests it's 2.2.0.3... hmmm 1197526462 M * otaku42 which means that the LBACK_REMAP nflag won't work. doh. 1197526873 M * otaku42 latest backports kernel has 2.2.0.4... still not enough for LBACK_REMAP and friends. hmm. 1197528022 M * otaku42 ok, then a different approach: 1197528077 Q * quasisane Ping timeout: 480 seconds 1197528104 M * otaku42 i have several vservers that require access to the loopback interface. so far i've dealt with that by adding an "entry" to /etc/vservers//interfaces/x with ip=127.0.0.1, prefix=8 and dev=lo 1197528130 M * otaku42 this works, but as soon as i shut down one of these vservers lo disappears inside the other 1197528133 M * otaku42 erm 1197528149 M * otaku42 this works, but as soon as i shut down one of these vservers the loopback interface disappears inside the other vservers 1197528180 M * otaku42 i tried my luck with sysctl net.ipv4.conf.all.promote_secondaries=1 but that didn't help 1197528230 M * otaku42 i take it the way i make the loopback interface available for the vservers that require it is wrong. which rises the question for alternative/better ways. any hints? 1197528624 N * DoberMann[ZZZzzz] DoberMann 1197528839 J * besonen_mobile__ ~besonen_m@71-220-198-145.eugn.qwest.net 1197528847 Q * besonen_mobile_ Read error: Connection reset by peer 1197529228 J * sharkjaw ~gab@shell.ormset.no 1197529297 J * balbir ~balbir@59.145.136.1 1197529318 J * jan ~jan@212.103.134.208 1197529348 M * jan Is it possible that my server is crashing becouse I have set memory limits on it for guests? 1197530076 J * JonB ~NoSuchUse@kg1-98.kollegiegaarden.dk 1197530430 M * cehteh maybe 1197530521 M * jan Aren't those limits preventing the new programs to start? 1197530548 M * jan I don't know how it works 1197530732 Q * ntrs__ Quit: Leaving 1197531298 M * cehteh i dont know which limits you use, normal resource limits will limit single processes 1197531336 M * cehteh but iirc (i am not really up to date) vserver can limit the memory a single vserver might use at max and then it will disallow programs to start 1197531347 M * ard6 otaku42 : You have to add an option to prevent the ip to be added or to be removed. I think touching nodev would suffice. 1197531359 M * cehteh use this limit very graceful, only to catch worst cases 1197531389 M * ard6 otaku42 : But if you think you need it, your setting is with 99.9999% certainty wrong :-) 1197531423 M * ard6 otaku42 : access to 127.0.0.1 will be rewritten to the first IP on your vserver 1197532171 N * jan yang 1197532908 N * DoberMann DoberMann[PullA] 1197533193 M * amaxik hi all 1197533227 Q * zLinux Ping timeout: 480 seconds 1197533238 M * amaxik daniel_hozac do you have any usable example guest settings for mem/cpu limits 1197533245 M * daniel_hozac sure. 1197533257 M * amaxik dont know what is optimal to limit my simple guests... 1197533257 M * daniel_hozac just set rlimits/rss.{hard,soft} and sched/* 1197533270 M * amaxik please, show ? ^) if not hard. 1197533278 M * daniel_hozac the settings depend entirely on what you want to do... 1197533291 M * amaxik yes. but need to be started from ... 1197533304 M * amaxik not from scratch. ) 1197533323 M * daniel_hozac run it unlimited for a while and check /proc/virtual//{limit,sched}... 1197533400 M * amaxik ah! 1197533403 M * amaxik one more bug.. 1197533432 M * amaxik for example.. having guest w/o dlimits. its running. setting up dlimit. restarting guest. it hangs 1197533446 M * daniel_hozac no, it's calculating the usage. 1197533447 M * amaxik with messages about "cant remove dlimit@ 1197533466 M * daniel_hozac run ps faux in another shell and notice the vdu process... 1197533513 M * amaxik also.. why we have 2 / mounts in guests ? 1197533525 M * amaxik is it correct ? 1197533537 M * daniel_hozac because that's the way Linux shows it? 1197533577 M * amaxik well that simple isnt comfortable to see 2 / mount and fix some software 1197533586 M * amaxik monitoring software and etc.. 1197533643 M * yang cehteh: Well, I think the limit could be the cause of my server crashing, I can't find anything else in the logs... 1197533676 M * cehteh yang: normally such killing shoud go into the logs 1197533699 M * daniel_hozac yes, OOM-kills are logged. 1197533704 M * yang which file 1197533712 M * cehteh what did you tried? distribute memory acros servers? 1197533733 M * daniel_hozac dmesg 1197533735 M * yang I made memory RSS limit 1197533742 M * yang to 500MB 1197533756 M * yang originally there is 1Gb ram and 2Gb swap 1197533771 M * cehteh thats reasonable and harmless 1197533801 M * yang I don't claim this is the cause of my server being stopped, but I thinked of it as an option 1197533805 M * cehteh but if you limit virtual memory for no good reason you are doomed 1197533841 M * yang Well, I don't like the users on the server exceeding the limit with some program, so I have limited it 1197533843 M * otaku42 ard6: thanks for the hint. is the "will be rewritten" stuff true also for v2.2.0.x? as far as i understood that's a feature of 2.3.x 1197533946 M * amaxik daniel_hozac what is easy way to do reboot a guest inside guest ? 1197533956 M * daniel_hozac amaxik: reboot -f 1197533985 M * amaxik thanks, will try. 1197534005 M * amaxik im amazed of quality of VServer solution 1197534031 M * amaxik now using 4 servers 8cpus each.. amazing stability 1197534324 J * _marv ~marv@modemcable128.145-80-70.mc.videotron.ca 1197534565 J * pmjdebruijn ~pascal@62.133.206.55 1197534570 M * pmjdebruijn hi 1197534574 M * daniel_hozac hello 1197534590 M * pmjdebruijn vserver Test enter 1197534590 M * pmjdebruijn vlogin: openpty(): I/O error 1197534603 M * pmjdebruijn this happens with a single vserver 1197534606 M * pmjdebruijn what could it be? 1197534647 M * pmjdebruijn other vservers are just fine 1197534732 J * shuri ~shuri@64.235.209.226 1197534740 J * Alikus ~alikus@217.150.200.212 1197534841 M * pmjdebruijn oh CRAP 1197534849 M * pmjdebruijn corrupted filesystem... never mind 1197534865 M * otaku42 ard6: well, i just tried to ping 127.0.0.1 in a vserver that has no interface-entry for lo (as described above). result: doesn't work. 1197534889 M * otaku42 ard6: however, your hint regarding "nodev" did the trick for me, so i take this route. thanks for the hint! 1197534917 M * otaku42 daniel_hozac: and thanks to you for your help earlier. 1197534945 M * daniel_hozac otaku42: ping is inherently bad for testing that. 1197534948 M * daniel_hozac otaku42: use TCP/UDP. 1197534982 M * otaku42 daniel_hozac: hmmkay... 1197534999 M * _marv is it possible to get more than 16 ip's to a vserver? i saw a 64 ip patch but i was thinking more of a 255 ip thing, give a hole block to a vserver... 1197535107 M * daniel_hozac _marv: in 2.3, you can assign entire networks. 1197535125 M * daniel_hozac _marv: and there is no limit at all. 1197535133 M * _marv k 1197535136 M * _marv is it pretty stable? 1197535144 M * _marv i havent tried the dev version 1197535174 M * yang cehteh: which log file should i look at for the memory crash? 1197535198 M * cehteh syslog messages .. or just dmesg 1197535200 M * otaku42 daniel_hozac: hmm, indeed, you are right. seems i have to investigate why i had problems before with one vserver that did not see "lo". 1197535248 M * _marv does 2.3 run on 2.6.23.x? or still 2.6.22.x? 1197535499 M * daniel_hozac we're skipping 2.6.23 entirely. 1197535524 M * daniel_hozac i run the devel versions only. 1197535648 M * _marv k 1197535667 J * zLinux ~zLinux@88.213.17.232 1197535711 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1197536015 J * larsivi ~larsivi@85.221.53.194 1197536204 M * _marv dan, mmm can't find a doc on assigning an entire network... 1197536221 M * _marv i got 2.3 up & running and now the vserver has all the ip's :) 1197536258 M * daniel_hozac _marv: naddress --nid --add --mask ... 1197536378 J * WDTY ~chatzilla@88.159.1.100 1197536492 J * gebura ~gebura@77.192.186.197 1197537009 M * gebura hi 1197537136 M * yang cehteh: I located the error - here it is http://openpaste.org/en/4298/ 1197537300 N * DoberMann[PullA] DoberMann 1197537634 J * dna ~dna@139-231-dsl.kielnet.net 1197537794 M * yang and here is the dmesg log after reboot http://openpaste.org/en/4299/ 1197538335 A * WDTY *coughs* 1197538472 Q * _marv Ping timeout: 480 seconds 1197539092 Q * zLinux Ping timeout: 480 seconds 1197539122 J * _marv ~marv@modemcable128.145-80-70.mc.videotron.ca 1197539783 Q * JonB Ping timeout: 480 seconds 1197540075 J * zLinux ~zLinux@88.213.32.101 1197540601 M * yang cehteh: it was a damaged drive error 1197541039 J * mire ~mire@91-171-222-85.adsl.verat.net 1197542479 M * cehteh eh hi .. yes .. /me was away 1197542649 Q * WDTY Read error: Connection reset by peer 1197543045 N * Bertl_zZ Bertl 1197543052 M * Bertl morning folks! 1197543118 M * Bertl yang: your SATA controller or cabling is bad (maybe the drive itself) 1197543342 M * Bertl (not that it would be Linux-VServer related :) 1197543664 P * _marv Leaving 1197544476 J * JonB ~NoSuchUse@130.227.63.19 1197544852 Q * balbir Ping timeout: 480 seconds 1197544907 J * virtuoso ~s0t0na@ppp91-122-138-220.pppoe.avangard-dsl.ru 1197545310 Q * virtuoso_ Ping timeout: 480 seconds 1197545437 N * DoberMann DoberMann[PullA] 1197546307 M * Radiance goodmorning all 1197546307 Q * zLinux Ping timeout: 480 seconds 1197546330 M * Radiance yang, did you try to check with smartctl the drive status ? 1197546919 J * Infinito argos@201-11-172-121.gnace701.dsl.brasiltelecom.net.br 1197547027 J * zLinux ~zLinux@88.213.27.81 1197547132 M * amaxik daniel_hozac i got a strange thing... 1197547144 M * amaxik after configuring dlimits 1197547150 M * amaxik inside guest 1197547165 M * amaxik df -h show me 1197547166 M * amaxik rootfs 2.0G 172K 1.9G 1% / 1197547185 M * Bertl and? 1197547186 M * amaxik like only 172k has used from 2gb... how ? 1197547216 M * Bertl probably the 172k is what 'belongs' to that context 1197547216 M * amaxik total limit per guest is 2gb. chxid was done. 1197547217 Q * mire Ping timeout: 480 seconds 1197547238 M * amaxik i did chxid for this guest for /vservers/xid 1197547246 M * Bertl could also be that your dlimit cache is wrong 1197547306 M * amaxik how to detect/fix ? 1197547353 M * daniel_hozac lsxid /vservers//usr/bin 1197547369 M * daniel_hozac does it show files of a size greater than 172K belonging to that guest? 1197547448 M * amaxik all files are marked to this context. 1197547500 M * daniel_hozac and when did you change the tagging? 1197547520 M * amaxik svn2 /vservers/svn2/usr/bin/bison 1197547523 M * amaxik show me 1197547528 M * Bertl probably after setting the limits 1197547530 M * amaxik and this file is >200k 1197547543 M * amaxik i have changed tagging before and after 1197547552 M * amaxik after setting up limit 1197547575 M * daniel_hozac stop the guest, remove the cache. 1197547582 M * amaxik hm.. 1197547584 M * amaxik cache 1197547587 M * daniel_hozac start it up and let it recalculate the usage. 1197547608 M * Bertl could it be that the cache is not listed on http://www.nongnu.org/util-vserver/doc/conf/configuration.html (yet)? 1197547625 M * amaxik cache is empty 1197547630 M * daniel_hozac did you stop the guest? 1197547633 J * WDTY ~chatzilla@88.159.1.100 1197547637 M * amaxik is empty /var/cache/vservers/svn2/dlimits 1197547644 M * amaxik hm.. 1197547647 M * Bertl wb WDTY 1197547657 M * WDTY Merci :) 1197547659 M * amaxik stoppped. 1197547676 M * daniel_hozac Bertl: it's there. 1197547678 M * amaxik space_used=1856 1197547678 M * amaxik inodes_used=323 1197547681 M * WDTY f#cking m$ auto updates 1197547685 M * daniel_hozac remove the file. 1197547703 Q * Aiken Quit: Leaving 1197547704 M * amaxik ye.. restarting. 1197547708 M * WDTY grr,got myself a nice " autoreboot" 1197547757 M * Bertl daniel_hozac: the cache link is there, yes, but no description of the files there, or am I blind? 1197547760 M * amaxik hm.. sloooww 1197547819 M * amaxik its has started. checked df -h 1197547833 M * amaxik only 104k from 5gb 1197547868 M * amaxik 2.6.22-vs2.3.0.28-gentoo 1197547876 M * amaxik maybe something wrong with this ? 1197547878 M * Bertl util-vserver version? 1197547900 M * amaxik 0.30.214 1197547915 M * Bertl filesystem? 1197547932 M * amaxik ^ /dev/mapper/vg09-vservers on /vservers type ext3 (rw,tag) 1197548069 M * amaxik is this a good way to use tagging ? or better is lvm2 partitions ? 1197548097 M * Bertl no, tagging is fine for shared guests (dlimits) 1197548109 M * daniel_hozac Bertl: you're right, but the filenames are based on the xid and path. 1197548165 M * amaxik so i have something strange with dlimits.. whats the easy way to fix it ? 1197548170 M * Bertl just wanted to mention it, as I was looking them up at least twice and couldn't find them :) 1197548220 M * amaxik should I upgrade kernel? or what is better for getting dlimits worked? 1197548220 M * Bertl amaxik: double check that the xid is correct 1197548237 M * amaxik xid is correct, sure. 1197548303 M * Bertl could you upload the output from 'vserver-info - SYSINFO' to paste.linux-vserver.org? 1197548316 M * daniel_hozac amaxik: and you restarted it and things are still wrong? 1197548386 M * amaxik daniel_hozac yes. stopped. deleted. started. results are wrong 1197548426 M * amaxik Bertl http://paste.linux-vserver.org/10325 1197548479 M * amaxik has used kernel with 01_patch-2.6.22.9-vs2.3.0.28.diff and 02_delta-cow-fix17.diff 1197548482 M * amaxik nothing more 1197548502 M * Bertl when you started the guest, what does 'vdlimit --xid /vservers show? 1197548598 M * amaxik http://paste.linux-vserver.org/10326 1197548656 M * Bertl daniel_hozac: --xid understands the guest name in all cases? 1197548694 M * amaxik chxid marks well with a guest name.. 1197548738 M * amaxik -c ... assign the given context/vserver to the file(s) 1197548827 M * daniel_hozac Bertl: yes. 1197548839 M * Bertl excellent, didn't know that 1197548861 M * amaxik Fine. ) 1197548894 M * Bertl well, the kernel side limits show that you have 188 blocks used (in 55 files) 1197548918 M * Bertl daniel_hozac: is there a way to check what vdu does? 1197548919 M * amaxik yes. 1197549099 M * daniel_hozac what do you mean? 1197549117 M * Bertl some debug mode or so? 1197549141 M * Bertl it seems to me like the initial values are calculated wrong for amaxik 1197549170 M * amaxik yes.. is there a way to recalculate ? 1197549197 M * Bertl removing the cache and starting the guest will do that (so that's not the problem) 1197549205 M * amaxik did it 1197549209 M * amaxik doing again 1197549227 M * Bertl I think the vdu calculation is going wrong for some reason 1197549275 M * amaxik hm!!! 1197549281 M * amaxik i found a strange thing 1197549405 M * daniel_hozac no, there's no verbose mode... 1197549425 M * amaxik so.. 1197549433 M * amaxik i found whats happened. 1197549468 M * amaxik but how to protect myself from this.. how to display dlimits configured per guest ? 1197549483 M * Bertl so, what happened? 1197549497 M * amaxik was a wrong path. 1197549508 M * amaxik in dlimit in /etc in guest. 1197549508 M * Bertl wrong path where? 1197549521 M * amaxik here /etc/vservers/svn2/dlimits/0/directory 1197549540 M * amaxik it shows to different dir (not to /vservers/svn2) 1197549543 M * Bertl ah, i.c. so you were setting the wrong ones 1197549555 M * Bertl probably copied the guest config or so 1197549561 M * amaxik yes. and i cant see actice dlimits per guest 1197549564 M * amaxik yes yes 1197549568 M * amaxik copied. 1197549581 M * Bertl don't do that (without fixing all the symlinks) 1197549587 M * amaxik how to protect myself from this bugs and show what really dlimits was configured ? 1197549589 M * Bertl use util-vserver for that 1197549617 M * amaxik vdlimit ? 1197549628 M * amaxik ok will try 1197549636 M * Bertl I meant for the copy (and it would be vserver ... build) 1197549680 M * amaxik vserver copy is non fuctioning still 1197549709 M * Bertl vserver - clone is what you want, IIRC (or rsync) 1197549726 M * amaxik ye. but here was an etc problem. 1197549749 M * Bertl because you 'copied' the config manually and didn't fix the symlinks :) 1197549760 M * amaxik so, will be very good to get cat /proc/something/xid/dlimits to show 1197549769 M * daniel_hozac huh? 1197549774 M * daniel_hozac vdlimit already does that... 1197549819 M * amaxik in /etc/vservers/svn2/dlimits/0/directory i have a real path to /vservers/svn2 is it wrong ? 1197549836 M * amaxik should I use there a symlink ? 1197549845 M * daniel_hozac no. 1197549877 M * amaxik in this case im trying see dlimits for /vservers/svn2 but if configured other pat 1197549902 M * amaxik path.. i unable to see what is configured totally (a paths per xid) 1197549930 M * amaxik ok, need a do dobule-check configs 1197549936 M * amaxik all thanks ) 1197550318 M * Bertl np 1197550630 J * lilalinux ~plasma@dslb-084-058-248-133.pools.arcor-ip.net 1197551512 M * amaxik heh 1197551516 M * amaxik new issue ) 1197551519 M * amaxik vsched: non-numeric value specified for '--priority_bias' 1197551544 M * amaxik file /etc/vservers/svn2/schedule was made as in wiki 1197551610 M * amaxik ah 1197551613 M * amaxik i see 1197551616 M * amaxik old syntax 1197551737 M * Bertl daniel_hozac: how much efford would it be to utilize the CGROUP namespace (given that we unshare it?) for the scheduler? 1197551781 M * daniel_hozac i have to admit that i haven't looked at cgroup at all yet. 1197551802 M * daniel_hozac i'm not a huge fan of the filesystem based APIs. 1197551906 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1197551926 M * Bertl I can understand that ... 1197552309 M * Bertl could we add a flag to select the pid space somehow? 1197552345 M * Bertl i.e. I do not want to put it in the unshare mask by default (or well, I would like to avoid using it by default) 1197552410 M * daniel_hozac sure. 1197552462 N * DoberMann[PullA] DoberMann 1197554229 Q * yang Quit: leaving 1197554641 J * Punkie ~Punkie@goc.coolhousing.net 1197554651 M * Bertl wb Punkie! 1197554662 M * Punkie hi Bertl :) 1197554984 Q * Infinito Quit: Quitte 1197555391 J * mire ~mire@91-171-222-85.adsl.verat.net 1197555495 M * Bertl wb mire! 1197555615 Q * JonB Ping timeout: 480 seconds 1197555651 J * balbir ~balbir@122.167.182.219 1197555838 Q * pmenier Read error: Connection reset by peer 1197555869 J * doener_ ~doener@i577B9BE5.versanet.de 1197555884 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1197556286 Q * doener Ping timeout: 480 seconds 1197556530 Q * shuri Quit: Leaving 1197556842 Q * larsivi Quit: Konversation terminated! 1197556852 M * Bertl daniel_hozac: same goes for the entwork namespaces, once they are operable 1197556860 M * Bertl (flag/option) 1197556869 M * daniel_hozac yeah. 1197557484 M * the-me hello 1197557494 M * Bertl wb the-me! 1197557511 M * the-me we've got an serious issue with 2.6.22.14-vs2.3.0.29: 1197557528 M * Bertl how so? 1197557541 M * ard6 ipv6 routing? :-) 1197557561 M * the-me after a while (a few days) we get an kernel panic. the root is online and pingable, if I connect to any service with telnet (host and guests) they return absolutly nothing 1197557572 M * the-me here is the only thing we could crap after the last crash: http://img252.imageshack.us/my.php?image=blubbzg1.png 1197557657 M * Bertl looks like an ext3 issue .. the filesystem is okay? 1197557699 M * the-me it should be okay, it's a new server and a new system 1197557717 M * the-me hmp on this systems there are a few customers :) 1197557730 M * Bertl you should capture the 'upper' part of that oops (trace) 1197557745 M * the-me heh we tried, but we wasn't able for it 1197557746 M * Bertl without the trace information, we cannot do much 1197557756 M * the-me the system was completly stocked 1197557763 M * Bertl get a serial console or change the display resolution 1197557798 M * the-me hmm an ext3 filesystem with a softwareraid 1197557843 J * hparker ~hparker@linux.homershut.net 1197557846 M * Bertl works perfectly fine here, so not a problem per se 1197558104 Q * WDTY Quit: ChatZilla 0.9.79 [Firefox 2.0.0.11/2007112718] 1197559353 J * Julius ~julius@p57B25B94.dip.t-dialin.net 1197559609 Q * Julius Remote host closed the connection 1197559612 Q * sharkjaw Quit: Leaving 1197560046 J * Julius ~julius@p57B25B94.dip.t-dialin.net 1197560423 J * JonB ~NoSuchUse@192.38.8.25 1197561199 J * dowdle ~dowdle@scott.coe.montana.edu 1197561415 M * Bertl wb dowdle! 1197561423 A * dowdle grunts 1197561817 J * Infinito argos@201-11-172-121.gnace701.dsl.brasiltelecom.net.br 1197561828 M * Bertl wb Infinito! 1197562018 J * Abaddon abaddon@68-71.is.net.pl 1197562846 M * Bertl wb Abaddon! 1197562915 M * Abaddon hi Bertl :) 1197563009 Q * pmjdebruijn Quit: Ex-Chat 1197563934 Q * gebura Quit: Quitte 1197563971 Q * Alikus Remote host closed the connection 1197564184 J * Alikus ~alikus@217.150.200.212 1197564312 Q * Alikus Remote host closed the connection 1197564796 Q * JonB Quit: This computer has gone to sleep 1197565286 M * Bertl okay, off for now .. back later 1197565292 N * Bertl Bertl_oO 1197565536 Q * lilalinux Remote host closed the connection 1197565819 J * virtuozzo ~rpavlyuk@195.160.234.1 1197565901 P * virtuozzo 1197565919 J * bonbons ~bonbons@2001:960:7ab:0:20b:5dff:fec7:6b33 1197566492 J * JonB ~NoSuchUse@kg1-98.kollegiegaarden.dk 1197567179 Q * hparker Remote host closed the connection 1197567293 Q * JonB Quit: This computer has gone to sleep 1197568032 N * DoberMann DoberMann[PullA] 1197568244 J * JonB ~NoSuchUse@kg1-98.kollegiegaarden.dk 1197568334 J * ViRUS ~mp@p57A6CE6D.dip.t-dialin.net 1197568675 J * shuri ~shuri@64.235.209.226 1197569063 Q * bonbons Quit: Leaving 1197569300 Q * pmenier Quit: Konversation terminated! 1197569418 J * paralizer ~paralizer@79.125.192.60 1197569444 Q * paralizer 1197569565 J * paralizer ~paralizer@79.125.192.60 1197569597 M * paralizer Hello. 1197569607 M * paralizer Anybody here? 1197569643 M * daniel_hozac hi 1197569663 M * paralizer I need some help. 1197569671 M * paralizer I have installed vserver virtualization. 1197569683 M * paralizer Ane now I have some problems with the available disk space. 1197569713 M * paralizer Filesystem Size Used Avail Use% Mounted on 1197569714 M * paralizer /dev/hdv1 110G 99G 2.8G 98% / 1197569727 M * paralizer I says 2.8GB available. 1197569748 M * paralizer But according to Size and Used it should be 11 GB? 1197569763 M * paralizer What is wrong? 1197569765 M * daniel_hozac that doesn't sound Linux-VServer related... 1197569782 M * daniel_hozac check the reserved percentage. 1197569811 M * paralizer Well it is a virtual server running on linux-.vservers 1197569813 M * paralizer Can you please tell me how? 1197569837 M * daniel_hozac it depends on the filesystem. 1197569856 M * daniel_hozac it's a general Linux thing, nothing specific to Linux-VServer... 1197569860 M * paralizer It is ext3 1197569886 M * paralizer Also, when I execute the chxid -c xxx -Rx xxxxx command i get the following error. 1197569888 M * paralizer lstat(): No such file or directory 1197569903 M * paralizer I think that the problem is somehow related to this error. 1197569920 M * paralizer But I can not find anything specific on the internet. 1197569932 M * daniel_hozac dumpe2fs 1197569956 M * daniel_hozac Reserved block count is what you're interested in... 1197570051 M * paralizer Ok, but where the hell are thos 11 GB? 1197570053 M * paralizer :) 1197570057 M * paralizer *those 1197570085 M * daniel_hozac what does that number * block size give you? 1197570099 M * paralizer It is still counting. 1197570116 M * daniel_hozac hmm? 1197570139 M * paralizer these are the lase 20 lines. 1197570141 M * paralizer Group 1364: (Blocks 44695552-44716918) 1197570141 M * paralizer Block bitmap at 44695552 (+0), Inode bitmap at 44695553 (+1) 1197570141 M * paralizer Inode table at 44695554-44696065 (+2) 1197570141 M * paralizer 2 free blocks, 11926 free inodes, 203 directories 1197570141 M * paralizer Free blocks: 44707250, 44708471 1197570142 M * paralizer Free inodes: 22347788, 22347791, 22347803, 22347831, 22347836, 22347838, 22347843, 22347894, 22350503-22350507, 22350512, 22350539, 22351099, 22351107, 22351114, 22351127-22351128, 22351144-22351157, 22351159-22351160, 22351164-22351166, 22351169-22351174, 22351176-22351177, 22351196-22351202, 22351204, 22351208, 22351210-22351211, 22351215-22351216, 22351234, 22351247-22351250, 22351279-22351281, 22351308, 22351312-22351317, 22351660-22351661, 223 1197570147 M * paralizer 51663, 22351674, 22351680-22351681, 22351683-22351687, 22351690-22351697, 22351699-22351700, 22351705, 22351707-22351713, 22351715-22351716, 22351719-22351721, 22351748-22351845, 22351847-22351848, 22351850-22351851, 22351853-22351857, 22351859-22351877, 22351879-22351885, 22351891-22351916, 22351923-22351952, 22351989, 22351993, 22351995-22352053, 22352062, 22352064, 22352068-22352135, 22352137-22352309, 22352315, 22352331, 22352360-22352625, 223526 1197570152 M * paralizer 45-22353026, 22353029, 22353031-22353123, 22353125-22353142, 22353148-22353149, 22353151-22353152, 22353154-22353155, 22353157-22353159, 22353161-22353162, 22353178, 22353180-22353182, 22353184-22353191, 22353193, 22353195, 22353199-22353201, 22353203, 22353211, 22353217-22353218, 22353226, 22353235, 22353241-22353262, 22353264-22353423, 22353425, 22353427-22353431, 22353437-22353445, 22353447-22353448, 22353451, 22353453, 22353456, 22353458-22353469 1197570157 M * paralizer , 22353476-22353477, 22353489-22353490, 22353496, 22353498-22353500, 22353549-22353607, 22353613-22353615, 22353795, 22353861-22353862, 22353864-22353881, 22353884-22353885, 22353888-22353889, 22353894, 22353899-22354080, 22354088-22354092, 22354111-22354155, 22354170-22354171, 22354173-22354180, 22354183-22354184, 22354186-22354470, 22354472-22364160 1197570164 M * paralizer *last 1197570165 M * daniel_hozac please use paste.linux-vserver.org for anything longer than 3 lines. 1197570182 M * daniel_hozac and you might want to grep for Reserved... 1197570190 M * paralizer Ok, sorry, I'm new here. 1197570217 M * paralizer dumpe2fs 1.35 (28-Feb-2004) 1197570218 M * paralizer Reserved block count: 2235845 1197570218 M * paralizer Reserved blocks uid: 0 (user root) 1197570218 M * paralizer Reserved blocks gid: 0 (group root) 1197570230 Q * amaxik Ping timeout: 480 seconds 1197570239 M * daniel_hozac and what block size? 1197570290 M * paralizer I do not know. When I executed ' dumpe2fs /dev/sda7 | grep Reserved' 1197570299 M * paralizer I get just those results. 1197570305 M * daniel_hozac so... grep for Block size... 1197570316 M * paralizer Ok. 1197570351 M * paralizer dumpe2fs 1.35 (28-Feb-2004) 1197570351 M * paralizer Block size: 4096 1197570372 M * daniel_hozac and... what does 4096 * 2235845 equal? 1197570408 M * paralizer 9158021120 1197570433 M * daniel_hozac so that would account for all the space, no? 1197570495 M * paralizer I'm sorry but I do not understand you well. 1197570549 M * daniel_hozac 110 * 1024 * 1024 * 1024 =~ 9158021120 + 2.8 * 1024 * 1024 * 1024 1197570774 M * paralizer Still, not so clear to me. So you want to say that this is not linux-verver related probem, but it is a problem with the OS itself? 1197570852 M * daniel_hozac there is no problem... 1197570895 M * daniel_hozac you have reserved 8.5 GiB for root. 1197570916 M * paralizer Ahaaaaa.... 1197570925 M * paralizer I used to have this problem in the past.... 1197570949 M * paralizer And how can I assignt this "lost" disk space to my /vservers partition? 1197570955 M * paralizer *assign 1197570966 M * paralizer You do not have to explain me. 1197570976 M * daniel_hozac what? 1197570977 M * paralizer You can just point me to some tutorial so something similar... 1197571036 M * paralizer How can I 'unreserved' those 8,5 GB? 1197571049 M * daniel_hozac man tune2fs 1197571049 M * paralizer I hope you can understand me.... 1197571055 M * paralizer Ok, thanks. 1197571056 M * paralizer :) 1197571594 Q * Abaddon Quit: leaving 1197571878 J * zbyniu_ ~zbyniu@host13-188.crowley.pl 1197571897 Q * zbyniu_ 1197572150 Q * Punkie Quit: Odcházím 1197572351 M * arachnist http://www.thelocal.se/9398.html <| lol 1197572360 Q * eviljonny Ping timeout: 480 seconds 1197572653 J * quasisane ~sanep@c-76-118-191-64.hsd1.nh.comcast.net 1197573103 Q * paralizer Quit: Leaving 1197573321 J * eviljonny ~eviljonny@loki.eviljonnys.com 1197573663 Q * JonB Quit: This computer has gone to sleep 1197573751 Q * Julius Quit: Verlassend 1197574428 J * JonB ~NoSuchUse@kg1-98.kollegiegaarden.dk 1197576342 J * ekeller ~chatzilla@ekellerpc.ee.Princeton.EDU 1197576399 M * daniel_hozac hi ekeller. sorry i haven't replied yet, i haven't gotten around to testing it... 1197576429 M * ekeller It's ok, I was just hoping you could drop me a few nuggets of hints of where the problem might be 1197576435 M * daniel_hozac though i'm somewhat certain it's unrelated. 1197576495 M * daniel_hozac feel free to test it yourself though, task_running_tick's rt_task branch is what i was curious about. 1197576530 M * ekeller I checked rt_prio(p), and it returns false. If that's what you're referring to. 1197576587 M * daniel_hozac task_running_tick is not executed at all? 1197576624 M * ekeller That I don't know. 1197576720 M * ekeller Perhaps I'm misunderstanding... is there an actual function called task_running_tick, or are you referring to scheduler_tick? 1197576758 M * ekeller In scheduler_tick there's a check if (rt_task(p)) {...}, and which I was saying that returns false. 1197576764 J * jmcaricand ~jmcarican@d77-218-122-10.cust.tele2.fr 1197576785 M * daniel_hozac well, it's task_running_tick in my tree. 1197576790 M * daniel_hozac that's 2.6.22-vs2.3 though. 1197576822 M * ekeller Ah, ok. My question is in scheduler_tick vx_need_resched gets called, but when tokens gets down to 0, the NEED_RESCHED (or whatever it's called) gets set. 1197576823 Q * jmcaricand 1197576862 M * ekeller But then in schedule(), the vx_schedule function gets called, which actually puts the task on hold if it runs out of tokens. 1197576905 M * ekeller What's the reason the task isn't put on hold immediately? 1197576925 M * ekeller or would this not matter? 1197577257 M * daniel_hozac you need to hold the other tasks in the context too. 1197577631 M * ekeller Do you have any other theories? My theory was that a task runs out of tokens, but it's a while before schedule() gets called to put on hold, but since it can't go into negative tokens, it quickly gets refilled. 1197577690 Q * dowdle Quit: Konversation terminated! 1197577760 J * Aiken ~james@ppp121-45-246-228.lns2.bne4.internode.on.net 1197577775 J * dowdle ~dowdle@scott.coe.montana.edu 1197578005 M * daniel_hozac that doesn't seem plausible to me. 1197578015 M * daniel_hozac especially not in the infinite loop case. 1197578105 M * ekeller Where else can I look? I see the tokens get decreased to 0 in /proc/virtual/777/sched. 1197578135 M * daniel_hozac so it's just the holding that doesn't work? 1197578249 M * ekeller I don't really know, is that the only thing that constrains the task to a certain CPU usage? If the task just gets rescheduled, if it's one of only a few things running (and it's fairly high priority), wouldn't it get to run again fairly quickly? 1197578302 Q * Infinito Quit: Quitte 1197578400 M * daniel_hozac if you set the sched_pause flag, does that successfully keep it from running? 1197578693 M * ekeller I'm not certain. I will check. 1197580109 M * ekeller I may have done something wrong... in vx_need_resched, when tokens got down to 0, I set the VXF_SCHED_PAUSE flag, and the thread does stop running. Though it never becomes unpaused, so I can't tell if it pauses immediately of after a while. 1197580165 M * daniel_hozac you can unset it with vattribute --set --xid --flag ~sched_pause 1197580499 M * ekeller That did some weird things to my computer. But I think the process is just staying paused (perhaps I unpause it and then it quickly pauses again). 1197580547 M * daniel_hozac hmm? 1197580551 M * daniel_hozac what kind of weird things? 1197580617 M * ekeller It was as if I hit enter hundreds of times. It's not doing it anymore and it could be unrelated. 1197580645 M * daniel_hozac you should undo your previous modification. 1197580666 M * daniel_hozac then setting sched_pause should pause it, until you unset it. 1197580708 M * ekeller What's the difference between pause and hold? 1197580723 M * daniel_hozac hold is what happens when it runs out of tokens. 1197580731 M * daniel_hozac pause is the manual version. 1197580841 M * ekeller Ok. Anyway, I just wanted to check in with you to see if you had any ideas. This is killing me that I can't figure it out, not to mention my project depends on getting it working. If you have any further thoughts, please let me know. 1197580945 M * daniel_hozac you'll want to wait for Bertl_oO to get back. he knows the scheduler. 1197581014 M * ekeller Ok, I'll try him later 1197581022 J * bonbons ~bonbons@2001:960:7ab:0:20b:5dff:fec7:6b33 1197581196 P * ekeller User is away. 1197581567 J * Abaddon abaddon@68-71.is.net.pl 1197582109 N * mattzerah`afk mattzerah 1197582483 J * edeschen ~edeschen@proxy-sjc-2.cisco.com 1197582617 M * edeschen Hi daniel, its me again. as you suggested, putting "half -f" in /etc/init.d/halt works so that vserver stop does not hang (I just had to figure out to remove the killall5 from that script). 1197582681 Q * shuri Quit: Leaving 1197582787 M * edeschen I also figured out that the runlevel commands was returning "unknown" in the vserver - the issue was that the default script in the vserver /etc/rc.d/rc.sysinit.vserver removes all /var/run, which is removing utmp which contains the runlevel information 1197582831 M * edeschen I still have a question, can anybody tell me why "init" in the vserver receives runlevel 6 when vserver stop is invoked 1197582881 M * daniel_hozac because there's no difference between halt and reboot. 1197582895 M * edeschen hmmm? 1197582932 M * edeschen ok, so runlevel 0 is simply never used in vserver? 1197582942 M * daniel_hozac right. 1197582982 M * edeschen ok, quick question, is vserver-stop going through /dev/initctl to notify the init running in the vserver? 1197583016 M * daniel_hozac no. 1197583025 M * edeschen I tried to look at the vserver scripting, I see a lot of pipe but couldn't see anything related to initctl 1197583031 M * daniel_hozac it sends SIGINT to init. 1197583044 M * edeschen ah ok 1197583066 M * edeschen and then init in vserver figures out that is to reboot hence runlevel 6? 1197583075 M * daniel_hozac yes. 1197583078 M * edeschen based on that SIGINT reception 1197583082 M * edeschen ok that makes sense 1197583086 M * daniel_hozac SIGINT is the ctrl-alt-del signal. 1197583110 M * edeschen ok cool 1197583125 M * edeschen thanx for all the help and the explanation daniel, really papreciated 1197583131 M * edeschen appreciated 1197583142 M * daniel_hozac you're welcome! 1197583462 J * ekeller ~chatzilla@pool-71-168-235-192.cmdnnj.fios.verizon.net 1197583515 N * mattzerah mattzerah`away 1197583956 P * ekeller 1197584175 J * Infinito argos@200-103-111-159.gnace701.dsl.brasiltelecom.net.br 1197584754 M * Bertl_oO off to bed now .. have a good one everyone! 1197584759 N * Bertl_oO Bertl_zZ 1197584839 Q * bonbons Quit: Leaving 1197585084 J * pflanze ~chris@vslab-dhcp-18.inf.ethz.ch 1197585090 M * pflanze Hello. 1197585126 M * pflanze What exactly is "make distcheck" meant to do? 1197585147 M * pflanze BTW util-vserver-0.30.214 fails in it. 1197585591 N * DoberMann[PullA] DoberMann 1197585754 M * pflanze (And some wondering: the barrier attribute is lost upon unmounting/mounting the filesystem. Ok, I've been using a startup script which sets the barrier flag after each boot, but still wondering. What if I umount+mount manually after boot?..) 1197586269 P * edeschen 1197586475 Q * JonB Quit: This computer has gone to sleep 1197586618 Q * ViRUS Quit: Leaving 1197586789 Q * dna Quit: Verlassend 1197588680 N * DoberMann DoberMann[ZZZzzz]