1258329692 Q * Piet Remote host closed the connection 1258329764 J * Piet ~piet@04ZAACA6J.tor-irc.dnsbl.oftc.net 1258330796 Q * scientes Remote host closed the connection 1258331191 M * biz Bertl: when using initstyle=plain, how is the init process killed on "vserver foo stop"? (Let's assume SysV-Init first so I know where to look for) 1258331230 M * Bertl it is sent a signal first, then, after some timeout, killed via vkill 1258331618 Q * AmokPaule Quit: Nettalk6 - www.ntalk.de 1258331628 M * biz ah 1258331630 M * biz INITKILL_SEQ=( 15 5 9 ) 1258331656 M * biz hrm... weird, let's see if I can get upstart to stop with one of them 1258331668 M * biz because I end up with 9 / SIGKILL 1258331678 M * biz ;( 1258331849 M * biz $ chcontext --xid 108 -- kill -15 1 1258331857 M * biz $ chcontext --xid 108 -- kill -5 1 1258331868 M * biz $ chcontext --xid 108 -- ps aux 1258331869 M * biz USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND 1258331869 M * biz root 1 0.0 0.0 19052 1324 ? S 01:34 0:00 /sbin/init 1258331869 M * biz root 17748 0.0 0.0 15164 1120 pts/2 R+ 01:37 0:00 ps aux 1258331877 M * biz no reaction :-( 1258331891 M * Bertl man init 1258331918 M * Bertl SIGINT Normally the kernel sends this signal to init when CTRL-ALT-DEL is pressed 1258332021 M * biz ah, so for sysv init you rely on the configuration of the ctrl-alt-del action? 1258332035 M * biz because for upstart I have removed that, I'll have a look what's in there 1258332219 M * biz exec shutdown -r now "Control-Alt-Delete pressed" 1258332385 M * biz err 1258332395 M * biz 15 = TERM, 5 = TRAP 1258332722 M * Bertl the SIGINT should be sent first (simulating the CTRL-ALT-DEL) if the guest doesn't act on that, it will be killed off 1258333095 M * biz hmm, ok. I've put the event handler back in there 1258333124 M * biz $ chcontext --xid 108 -- kill -SIGINT 1 1258333129 M * biz then following it with 1258333135 M * biz $ chcontext --xid 108 -- ps aux 1258333149 M * biz showed that it ran shutdown -r now "..." 1258333157 M * biz and entered runlevel 6 1258333165 M * biz all services were killed 1258333176 M * biz but not init itself :-/ 1258333191 M * biz so it still remains to be killed afterwards with SIGKILL 1258333194 M * Bertl sounds good, the last action should be a 'halt' or 'reboot' issued from within 1258333231 M * Bertl if that happens, the helper on the host will eliminate the init/guest 1258333263 M * biz yeah, it does. however, I'm trying to get rid of: 1258333266 M * biz A timeout occured while waiting for the vserver to finish and it will 1258333266 M * biz be killed by sending a SIGKILL signal. The following process list 1258333266 M * biz might be useful for finding out the reason of this behavior: 1258333281 M * biz in the list is only /sbin/init left 1258333288 M * biz which is then SIGKILLED :) 1258333303 M * Bertl the timeout should not trigger if reboot/halt is issued 1258333316 M * Bertl (as it happens before the timeout :) 1258333375 M * biz ok, the /etc/rc6.d/K99reboot is triggered by the runlevel change, and it executes: 1258333391 M * biz reboot -d -f -i 1258333436 M * biz When called with --force or when in runlevel 0 or 6, this tool invokes the reboot(2) system call itself and directly reboots the system. 1258333467 M * Bertl yep, otherwise it just notifies init 1258333472 M * biz and this call within a guest should kill init? 1258333486 M * Bertl with a properly setup helper, yes 1258333507 M * biz hrmm. Looks like I'm missing this one :D 1258333710 M * Bertl should be configured by the util-vserver runlevel scripts (on the host) 1258333760 M * biz well I've switched from default (sysv) to plain by echo plain > /etc/vserver/foo/apps/init/style 1258333777 M * biz the only thing I did in addition was echo 3 > .../apps/init/runlevel 1258333820 M * Bertl should be fine, what does '/proc/sys/kernel/vshelper' contain? 1258333838 M * biz /sbin/vshelper 1258333990 M * Bertl and that one exists and is part of your current util-vserver install, yes? 1258334041 M * biz /sbin/vshelper -> /usr/lib/util-vserver/vshelper 1258334054 M * biz yes, it's in the util-vserver package 1258334095 M * Bertl assuming that upstart calls halt/restart (kernel side) that helper should activate and kill off the guest 1258334124 M * Bertl you might replace it with a wrapper script to log activation (just to make sure) 1258334302 M * biz I did, and it is not called 1258334345 M * biz I'll try to force a reboot in the context, let's see if init calls reboot(2) 1258334358 M * Bertl do you have Linux-VServer debugging enabled in the kernel? if so you can get an info about the helper delegation 1258334424 M * biz nope, that old system (with 2.6.22) is in production, that's also why I couldn't upgrade it easily ;) 1258334598 M * biz $ chcontext --xid 108 -- reboot -f 1258334605 M * biz vshelper called with args: restart 108 1258334647 M * biz hrmmm... wtf, why doesn't rc6.d/reboot trigger this 1258334650 M * biz *searching* 1258335068 M * biz ok, I've got it. Now "vserver foo stop" finally executes /etc/rc6.d/S90reboot, which does issue a reboot(2) call 1258335087 M * biz and the vshelper is called like "/sbin/vshelper restart 108" 1258335101 M * Bertl good, I presume that also takes care of the guest 1258335135 M * biz too bad it doesn't, I still encounter the timeout and it get's SIGKILLed 1258335144 M * biz ah wait. 1258335153 M * biz :)) 1258335213 M * biz indeed. It does! :-) 1258335232 M * Bertl great! 1258335247 M * biz thanks very much again, Bertl :) 1258335331 M * biz However, I'll have to fix that wiki entry 1258335395 M * biz http://linux-vserver.org/index.php?title=Upstart_issues&diff=prev&oldid=4071 1258335410 M * biz don't delete control-alt-delete.conf :) 1258335419 M * Bertl yeah, please do that :) 1258335424 M * Bertl and, you're welcome! 1258335639 M * biz Bertl: You've just a received a donation... to get some beer :D 1258335749 M * Bertl thanks! 1258335791 M * biz bye 1258335810 M * Bertl cya! 1258336312 J * Hurga ~hurga@dslb-088-078-048-112.pools.arcor-ip.net 1258336430 M * Hurga Hi. I just noticed a strange problem... after migrating my vserver guests to a new machine, the pie charts of my webalizer stats are missing colour. I don't get it. 1258336455 M * Bertl yeah, that indeed sounds strange :) 1258336478 M * Hurga happened in several guests, and I didn't touch at all at least two of them... 1258336504 M * Bertl what are the involved kernel/patch/util-vserver versions? 1258336538 M * Hurga The new root server is a 64bit machine, guests are 32 bit, old machine was 32bit,t oo., This is about the only difference which i can think of which *might* matter, 1258336606 M * Hurga 2.6.27.35-vs2.3.0.36.6 and util-vserver-0.30.216-pre2849 1258336658 M * Bertl that's the one you moved from? 1258336691 M * Hurga that's current. old one.. uh. Hang on 1258336782 M * Hurga util-vserver 0.30.211-1uk 1258336820 M * Hurga kernel was 2.6.17-11 but i have to dig for the vserver patch version. 1258336853 M * Bertl well, a bunch of things _might_ have changed there including but not limited to: 1258336869 M * Bertl - personality (you might need one for your 32bit guests) 1258336884 M * Bertl - localhost (127.0.0.1) handling 1258336892 M * Bertl - single_ip special casing 1258336908 M * Hurga 2.6.17-11.33vs2.0.2.1-server was the exact version BTW 1258336954 M * Hurga personality was what I was thinking of, too, heard you mention it earlier. I'll try that... 1258336961 M * Hurga it's weird still. 1258336988 M * Bertl to get backwards compatibility, set the personality to a 32bit one, and specify the one and only IP (for single IP guests) as lback address 1258337097 A * Hurga nods. "I think that the colour handling of gd lib has something to do with my loopback IP is really far-fetched, though. :) 1258337200 M * Bertl well, I've heard (and seen) stranger things ... 1258337217 M * Hurga ook... 1258337229 A * Hurga will keep that in mind. 1258337245 M * Hurga then again, that's why I'm asking here :) 1258337317 M * smash hey bertl / rest :) 1258337408 M * Hurga BTW, is it possible to check from inside the guest what personality is running? 1258337452 M * Bertl good question, I'd say yes 1258337630 M * Hurga got colour again with PER_LINUX32_3GB. 1258337649 M * Bertl great! 1258337667 M * Hurga Indeed. Thanks! 1258337674 M * Bertl you're welcome! 1258337734 M * biz I've added the upstart with current stable issues at the end of: http://linux-vserver.org/Upstart_issues 1258337755 M * Bertl excellent! 1258339481 J * infowolfe ~infowolfe@c-71-236-152-35.hsd1.or.comcast.net 1258339705 M * Bertl off to bed now ... have a good one everyone! 1258339712 N * Bertl Bertl_zZ 1258339775 A * Hurga gets some sleep, too. 1258339783 Q * Hurga Remote host closed the connection 1258340435 J * derjohn_foo ~aj@e180192039.adsl.alicedsl.de 1258340831 J * hparker ~hparker@208.4.188.201 1258340863 Q * derjohn_mob Ping timeout: 480 seconds 1258341831 Q * snooze Ping timeout: 480 seconds 1258342293 J * aj__ ~aj@e180192059.adsl.alicedsl.de 1258342667 Q * derjohn_foo Ping timeout: 480 seconds 1258343178 Q * daniel_hozac Ping timeout: 480 seconds 1258343895 J * saulus_ ~saulus@c193180.adsl.hansenet.de 1258344306 Q * SauLus Ping timeout: 480 seconds 1258344311 N * saulus_ SauLus 1258344670 J * daniel_hozac ~daniel@c-6c3771d5.08-230-73746f22.cust.bredbandsbolaget.se 1258344947 Q * aj__ Ping timeout: 480 seconds 1258345816 J * snooze ~o@1-1-4-40a.gkp.gbg.bostream.se 1258348299 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1258355320 J * aj__ ~aj@tmo-101-25.customers.d1-online.com 1258355830 Q * thierryp Remote host closed the connection 1258356425 J * sharkjaw ~gab@90.149.121.45 1258358062 J * davidkarban ~david@199.123.broadband11.iol.cz 1258359812 J * thierryp ~thierry@zankai.inria.fr 1258361128 J * fleischergesell ~fleischer@dslb-088-076-048-058.pools.arcor-ip.net 1258361245 J * friendly ~friendly@ppp118-209-136-134.lns20.mel6.internode.on.net 1258362694 Q * aj__ Ping timeout: 480 seconds 1258364873 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1258367048 Q * friendly Quit: Leaving. 1258367112 J * AmokPaule ~amokpaule@brsg-4dbbb2c3.pool.mediaWays.net 1258367223 J * aj__ ~aj@213.238.45.2 1258370461 Q * uva Quit: Leaving 1258370883 Q * renihs Quit: narf 1258371130 J * uva bno@114-45-227-25.dynamic.hinet.net 1258371879 M * fleischergesell whats the best approach for this: i have 3 guests, 2 of them use mount bind to make a dir from guest 1 available inside guest 2 and 3 - how do i setup quota so that it "works" for all three guests? 1258374757 Q * Piet Remote host closed the connection 1258374826 J * Piet ~piet@04ZAACBK7.tor-irc.dnsbl.oftc.net 1258375337 N * Bertl_zZ Bertl 1258375342 M * Bertl morning folks! 1258375357 M * fleischergesell hi Bertl 1258375382 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258375443 M * Bertl if you mean user/group quota, then you are out of luck for the shared partition, currently there is no context aware quota support in Linux-VServer, but of course, you can use quota (non context aware) on the shared partition or apply disk limits to it (when the filesystem is tagged) 1258375474 M * fleischergesell well, disk limits is kinda useless in my case unfortunately 1258375511 M * fleischergesell and what do you mean by non context aware quotas? 1258375537 M * fleischergesell just setting up usrquota 3 times on each guest? 1258375544 M * Bertl quota is based on uid/gid 1258375553 M * fleischergesell sure 1258375561 M * fleischergesell all users are present on all 3 guests 1258375571 M * Bertl i.e. a user with uid=100 will be handled the same in all guests 1258375604 M * fleischergesell oh, but you would setup the quota on the host then? 1258375634 M * Bertl if that is fine for you, then just setup quota in all three guests, but I wouldn't use bind mounts into one guest, I'd use a separate partition mounted into all three guests instead 1258375683 M * fleischergesell and you wouldnt mount with bind but how instead? 1258375697 M * Bertl just via the guest configs fstab 1258375722 M * fleischergesell well thats what i do - but as bind - didnt really work otherwise if i recall correctly 1258375787 M * Bertl guests use namespaces, so each guest can have a separate mount for that filesystem 1258375947 M * fleischergesell i see - did not really want to have a seperate partition, but seems to be the best approach 1258376068 M * Bertl it simplifies management when you 'share' something between guests, and it avoids the 'barrier problem' 1258376133 M * fleischergesell true 1258376175 M * fleischergesell gotta shrink my softraid array then, grml 1258376286 M * Bertl using lvm ontop of raidN is a good idea to keep it flexible :0 1258376315 M * fleischergesell hehe.. 1258376381 M * fleischergesell well, its alot easier to just recreate the partitions - backing up vservers is fortunately very easy.. 1258376492 M * Bertl correct 1258378563 Q * sharkjaw Remote host closed the connection 1258378658 J * blues ~blues@dqo85.neoplus.adsl.tpnet.pl 1258378778 Q * blues_ Ping timeout: 480 seconds 1258379234 J * marl_scot ~marl_scot@84.92.193.225 1258379442 M * marl_scot hi, have been playing around with virtualbox recently, and was wandering if anyone had ever installed it alongside vserver (2.2.0.6) ? is it posible and ifso is there anything i should look out for? I know it sounds like a daft thing to do, but have been running vserver for a number of years and would like to be able to have a couple of guests on my server running xwindows, setting up xwindows in a vserver requires opening up a few security holes fro 1258379443 M * marl_scot m what i remember, so thought a vbox install side by side would allow me to move vbox images from my desktop to my remote server 1258379496 M * Bertl Linux-VServer is quite unintrusive and works well with kvm/qemu/xen, so I presume virtualbox should be no different there 1258379500 M * marl_scot i would never replace vserver with vbox, has vbox requires to meny resources 1258379511 M * marl_scot thanks Bertl 1258379559 M * Bertl I use kvm for most of the kernel side Linux-VServer testing :) 1258379677 M * Bertl regarding X11, usually you do not need to run the X server inside a guest to have 'X11' apps or functionality 1258379714 M * Bertl but of course, if you are testing different X11 servers or configurations, you need the real thing :) 1258380229 M * marl_scot yup, going to try a few differnt ones, to find the one that uses the least amount of resources 1258383030 Q * BenG Read error: Connection reset by peer 1258383056 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258384013 Q * davidkarban Quit: Ex-Chat 1258384843 J * hparker ~hparker@208.4.188.201 1258385981 Q * thierryp Quit: ciao folks 1258386365 J * borgfish ~bla@141.12.67.138 1258386390 M * borgfish hi! since yesterday i have a problem with 1 vserver which i think might be vserver related 1258386403 M * borgfish too many open files for pam auth 1258386414 M * borgfish cat /proc/virtual/1074/limit 1258386434 M * borgfish Limit current min/max soft/hard hits 1258386440 M * borgfish FILES: 12167 0/ 15514 -1/ -1 0 1258386449 M * borgfish cat /proc/sys/fs/file-nr 1258386454 M * borgfish 14144 0 1202333 1258386457 M * borgfish any ideas ? 1258386535 M * Bertl check ulimits, the files might be limited there 1258386558 M * Bertl your per guest limits show no limit, so that's not the issue here 1258386558 Q * Mr_Smoke Read error: Connection reset by peer 1258386586 M * borgfish ulimit shows: unlimited 1258386593 M * borgfish its saslauthd using pam that fails 1258386610 J * Mr_Smoke ~smokey@layla.lecoyote.org 1258386613 M * borgfish for nearly 2 years it worked without an issue, now it failed twice in 2 days 1258386617 M * borgfish no changes to system 1258386718 M * Bertl check both the hard and soft limits for ulimit, and check as the user saslauthd is running as 1258386777 M * borgfish you mean i should check the limits imposed by vserver kernel ? i only have limited ram usage, saslauthd runs as root 1258386815 M * Bertl nah, there are no limits (at least to the number of files) for your guest 1258386839 M * Bertl i.e. it could be an ulimit you actually hit, or a global limit 1258386843 M * borgfish and if i look at the actual infos from vserver, the system had like 3000 less open files than its max. why would it still fail.. 1258386861 M * Bertl as I said, no limit from Linux-VServer side :) 1258386878 M * borgfish would vserver kernel log if it hit a limit ? 1258386878 M * Bertl -1/-1 means unlimited 1258386906 M * Bertl the hit count (last column) would increase 1258386923 M * borgfish ok 1258387040 J * dowdle ~dowdle@scott.coe.montana.edu 1258387195 J * geb ~geb@earth.gebura.eu.org 1258387869 Q * BenG Quit: I Leave 1258388488 N * DoberMann[ZZZzzz] DoberMann 1258389184 M * borgfish bertl thanks for your time, i will work arround with cronjob restarting saslauthd. wonder why it worked all the time... cya 1258389188 Q * borgfish 1258391222 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1258391614 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258391699 J * thierryp ~thierry@home.parmentelat.net 1258392162 J * ViRUS ~mp@p579B52A4.dip.t-dialin.net 1258393589 Q * gnuk Ping timeout: 480 seconds 1258393676 J * BenG_ ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258393696 Q * BenG_ 1258393766 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1258394008 Q * kir Quit: Leaving. 1258394864 Q * geb Quit: / 1258395303 J * BenG_ ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258395303 Q * BenG_ 1258395455 Q * BenG Quit: I Leave 1258396702 Q * balbir Ping timeout: 480 seconds 1258397366 J * balbir ~balbir@122.172.27.189 1258397829 Q * aj__ Ping timeout: 480 seconds 1258398250 Q * gnuk Quit: NoFeature 1258399400 J * BenG ~bengreen@94-169-110-10.cable.ubr22.aztw.blueyonder.co.uk 1258399538 J * geb ~geb@250.4.82-79.rev.gaoland.net 1258400562 Q * thierryp Quit: ciao folks 1258401064 J * aj__ ~aj@c195211.adsl.hansenet.de 1258401108 Q * BenG Quit: I Leave 1258403332 Q * balbir Ping timeout: 480 seconds 1258404085 J * balbir ~balbir@122.172.27.189 1258405929 J * hijacker_ ~hijacker@87-126-142-51.btc-net.bg 1258406017 Q * fleischergesell Ping timeout: 480 seconds 1258407501 Q * bonbons Quit: Leaving 1258407543 Q * hijacker_ Quit: Leaving 1258408546 Q * AmokPaule Quit: Nettalk6 - www.ntalk.de 1258410114 Q * aj__ Ping timeout: 480 seconds 1258410805 Q * dsoul Read error: Connection reset by peer 1258411387 J * AmokPaule ~amokpaule@brsg-4dbbb2c3.pool.mediaWays.net 1258412503 J * aland ~aland@81.95.52.68 1258412718 N * DoberMann DoberMann[ZZZzzz] 1258414799 Q * ViRUS Quit: If there is Artificial Intelligence, then there's bound to be some artificial stupidity. (Thomas Edison)