1150329618 J * doener ~doener@i5387C3CD.versanet.de
1150329832 Q * click Ping timeout: 480 seconds
1150329870 Q * doener_ Ping timeout: 480 seconds
1150330218 Q * afta Quit: 
1150331271 Q * trasher Ping timeout: 480 seconds
1150331337 J * trasher daniel@derichs.info
1150331697 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1150333240 Q * fwl Ping timeout: 480 seconds
1150333504 N * sarnold sars
1150334789 J * DarthVader ~Aniken@203.177.212.163
1150337992 J * DarthLord ~Aniken@203.177.212.165
1150338096 Q * DarthVader Ping timeout: 480 seconds
1150338558 Q * Hunger Read error: Connection reset by peer
1150338566 J * Hunger Hunger.hu@Hunger.hu
1150342939 Q * DarthLord Quit: Leaving
1150342955 J * DarthVader ~Aniken@203.177.212.165
1150344633 J * Zaki ~Zaki@88.213.19.115
1150345421 J * fwl ~f_@83-215-237-1.seek.stat.salzburg-online.at
1150345577 Q * fwl Quit: 
1150348462 J * morrigan_ morrigan@212.16.62.52
1150348463 J * nox_ ~nox@noxlux.de
1150348573 Q * morrigan Read error: Connection reset by peer
1150348598 Q * nox Ping timeout: 480 seconds
1150349146 M * mugwump so.  anyone got the SPs in new sunfires recording oops/panics ?
1150349381 M * mugwump or do I have to have a server connecting to them to log their console?
1150350511 J * ksf ~krazy_sys@202.80.169.52
1150354333 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at
1150354435 J * fwl ~f_@83-215-237-1.seek.stat.salzburg-online.at
1150354860 J * Viper0482 ~Viper0482@p5497759B.dip.t-dialin.net
1150355439 J * dna ~naucki@dialer-160-146.kielnet.net
1150356422 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1150357132 Q * h01ger Quit: h01ger
1150358224 Q * fwl Quit: This computer has gone to sleep
1150358261 Q * zkbrsnie Quit: 
1150358289 J * fwl ~f_@83-215-237-1.seek.stat.salzburg-online.at
1150358427 Q * Viper0482 Remote host closed the connection
1150359238 J * Viper0482 ~Viper0482@p5497759B.dip.t-dialin.net
1150359392 M * harry anyone?
1150359398 M * harry alive?
1150359525 Q * Aiken Ping timeout: 480 seconds
1150361460 M * harry to all: new grsec + vserver patch online
1150361855 J * genosid ~a@85.216.135.73
1150362092 J * h01ger ~holger@socket.layer-acht.org
1150362439 J * meandtheshell ~markus@85-124-175-238.dynamic.xdsl-line.inode.at
1150363806 Q * genosid Ping timeout: 480 seconds
1150364478 Q * ksf Ping timeout: 480 seconds
1150364959 M * orionpanda MugWump: are you using a SunFire T1000 or T2000? Also, why are you using Vserver instead of Solaris Zones?
1150364986 Q * shedi Quit: Leaving
1150365255 M * orionpanda harry: Where is the patch uploaded to? I can't find it under http://vserver.13thfloor.at/Experimental/
1150365274 M * harry http://ludit.kuleuven.be/software/vserver
1150365304 M * harry brb... x restart (new kde etc... :))
1150365324 M * orionpanda I'm using patch-2.6.16.20-vs2.1.1-rc22.diff. What is grsec for?
1150365329 M * orionpanda ok
1150365613 M * harry security
1150365686 M * orionpanda I'm reading through grsecurity.net now. Are there existing vserver exploits that grsec protects against? (In a hosting environment, for example?)
1150365755 J * jhaig ~jhaig@mantis.office.netline.net.uk
1150365804 M * sid3windr are there existing vserver exploits? :)
1150365948 M * jhaig I'm testing vserver with a view to using it for virtual dedicated servers.  I have just noticed that the load displayed by 'top' in a guest is the load for the host.  Is it possible to change this so that a guest only sees a value relevant to its own processes?  I am, for example, seeing a load of 1 because I am building a kernel on the host but the guest is not doing anything.
1150365969 M * sid3windr there are lots of options that can be virtualized
1150365974 M * sid3windr I don't know if loadaverage is one of them, but I think it is
1150365987 M * sid3windr I'm not sure how though, you need to set some flag somewhere I guess
1150366069 M * jhaig Is there some documentation about all the possible configuration options?  What I have found so far is the wiki, which seems to be a compilation of various peoples experimentation.  It is useful, but a comprehensive user guide would also be useful.
1150366081 M * sid3windr google "linux-vserver flower page" :)
1150366093 M * sid3windr hmm.
1150366096 M * jhaig Right, I've just found that now.
1150366098 M * sid3windr not the desired first hits :p
1150366098 M * sid3windr hehe
1150366106 M * sid3windr don't forget to set the stylesheet to something sane ;)
1150366122 M * harry there are no existing exploits for vserver afaik
1150366130 M * jhaig You mean something that doesn't have marijuana as a background?
1150366142 M * harry but grsec and pax provide an extra layer of protection by.... hell.. read the grsec page ;)
1150366146 A * harry had meeting now
1150366148 M * harry cya'll!
1150366154 M * sid3windr jhaig: for example ;)
1150366164 M * orionpanda harry: I'm reading the wiki. Grsec is amazing. Thanks for creating the patch!
1150366174 M * harry np
1150366175 M * harry ;
1150366176 M * harry :)
1150366235 M * jhaig OK, so how do I use an alternative style sheet in firefox?
1150366339 Q * Viper0482 Remote host closed the connection
1150366420 M * sid3windr in mozilla it's view -> use style
1150366422 M * sid3windr ff is prolly equivalent
1150366522 M * orionpanda What is the difference between the 2.1.1 and 2.0.2 vserver branches? I'm using 2.1.1
1150366528 M * jhaig Thanks.
1150366540 M * jhaig 2.1.1 is still in development
1150366554 M * orionpanda Yes, but functionally. Any differences?
1150366587 M * jhaig I don't know I'm afraid.  I've only been using it for a week and I'm still trying to find my way round.
1150366608 M * sid3windr the dev version has way more features afaik
1150366613 M * sid3windr more virtualization and the like
1150366618 M * sid3windr but possibly a bit less tested
1150366660 M * orionpanda so features like COW may be absent from 2.0.2? That's too bad because harry's grsec patch is for 2.0.2...
1150366673 M * sid3windr I think cow is only in 2.1 but I don't follow that too closely atm
1150366681 M * sid3windr just looking at 2.1 now because of ipv6 patches
1150366693 M * orionpanda yes, I'm using that, too. very useful
1150367045 M * jhaig Any idea when 2.1 may become stable?  If it is in the immediate future (next few weeks) I may start looking at that instead.  Otherwise I'll stay with 2.0.2
1150367157 M * orionpanda no idea. I haven't had any stability problems with it, however. Works great.
1150367682 M * kir Can anybody tell me whether one can use per-guest disk quotas with vserver?
1150367702 M * kir I mean, per-guest quotas set from the host system
1150367771 M * jhaig kir: I have just set that up myself, so yes you can.
1150367784 M * jhaig kir: http://linux-vserver.org/Disk+Limits
1150367799 M * jhaig I'm now wrestling with memory limits  :-/
1150368685 M * daniel_hozac 2.1 will _never_ be stable.
1150368697 M * daniel_hozac it's the development branch. it's supposed to be unstable.
1150368747 M * daniel_hozac (it's the same method the kernel uses(d?). x.y is stable, x.(y+1) is development)
1150368801 M * daniel_hozac however, as features get tested and verified to be stable, i guess they might be moved to the stable branch.
1150368826 M * sid3windr to 2.2 :)
1150368858 M * daniel_hozac well, some features (AIUI) will be moved to 2.0.
1150368951 M * sid3windr :)
1150368972 M * sid3windr damn, had to look up aiui ;)
1150369009 M * phedny Artificially Intelligent User Interface?
1150369087 M * orionpanda not aiui
1150369643 J * genosid ~a@chello085216135073.chello.sk
1150369863 M * jhaig Hmmm, I see it is possible to set the 'personality' of a context.  Is this like the personality setting of Eddie, the shipboard computer the Hitchhikers Guide to the Galaxy?
1150370022 Q * brc_ Quit: No windows for this server
1150370027 M * daniel_hozac more like linux32 for 32-bit guests on 64-bit hosts.
1150370052 M * jhaig Oh well, never mind  ;-)
1150372473 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1150372519 M * cdrx hi
1150372531 M * daniel_hozac hello
1150372943 J * _mcp ~hightower@wolk-project.de
1150372999 Q * mcp Read error: Connection reset by peer
1150373004 N * _mcp mcp
1150373010 Q * fwl iridium.oftc.net xenon.oftc.net
1150373010 Q * h01ger iridium.oftc.net xenon.oftc.net
1150373010 Q * ||Cobra|| iridium.oftc.net xenon.oftc.net
1150373010 Q * morrigan_ iridium.oftc.net xenon.oftc.net
1150373010 Q * doener iridium.oftc.net xenon.oftc.net
1150373010 Q * orionpanda iridium.oftc.net xenon.oftc.net
1150373010 Q * s0undt3ch iridium.oftc.net xenon.oftc.net
1150373010 Q * ddlp iridium.oftc.net xenon.oftc.net
1150373010 Q * Smutje iridium.oftc.net xenon.oftc.net
1150373010 Q * anonc iridium.oftc.net xenon.oftc.net
1150373010 Q * _Kara iridium.oftc.net xenon.oftc.net
1150373010 Q * starlein iridium.oftc.net xenon.oftc.net
1150373010 Q * independence iridium.oftc.net xenon.oftc.net
1150373010 Q * jake- iridium.oftc.net xenon.oftc.net
1150373010 Q * Adrinael iridium.oftc.net xenon.oftc.net
1150373010 Q * Snow-Man iridium.oftc.net xenon.oftc.net
1150373010 Q * weasel iridium.oftc.net xenon.oftc.net
1150373010 Q * ag- iridium.oftc.net xenon.oftc.net
1150373076 J * h01ger ~holger@socket.layer-acht.org
1150373076 J * fwl ~f_@83-215-237-1.seek.stat.salzburg-online.at
1150373076 J * ||Cobra|| ~cob@pc-csa01.science.uva.nl
1150373076 J * morrigan_ morrigan@212.16.62.52
1150373076 J * doener ~doener@i5387C3CD.versanet.de
1150373076 J * orionpanda orionpanda@netblock-66-245-252-180.dslextreme.com
1150373076 J * s0undt3ch ~s0undt3ch@bl7-240-252.dsl.telepac.pt
1150373076 J * ag- ag@caladan.roxor.cx
1150373076 J * Adrinael adrinael@hoasb-ff09dd00-79.dhcp.inet.fi
1150373076 J * jake- psybnc@murlocs.org
1150373076 J * independence independen@marcusson.no-ip.com
1150373076 J * starlein star@fo0bar.de
1150373076 J * Snow-Man ~sfrost@kenobi.snowman.net
1150373076 J * _Kara ~Kashira@wan14.ihredomainadresse.de
1150373076 J * anonc ~anonc@staffnet.internode.com.au
1150373076 J * weasel weasel@weasel.noc.oftc.net
1150373076 J * Smutje ~Smutje@xdsl-87-78-98-134.netcologne.de
1150373076 J * ddlp ~niko@dsl081-142-022.chi1.dsl.speakeasy.net
1150373212 J * click click@ti511110a080-2963.bb.online.no
1150374103 Q * meandtheshell Ping timeout: 480 seconds
1150374183 J * NetAsh ~NetAsh@62-212-199-5.elsis.lt
1150374205 P * NetAsh 
1150374297 J * meandtheshel1 ~markus@85-124-174-184.dynamic.xdsl-line.inode.at
1150374552 Q * DarthVader Quit: Leaving
1150375629 J * shedi ~siggi@213.190.108.50
1150375985 M * kir jhaig, thanks!
1150377611 M * harry orionpanda: reason i don't merge grsec and vserver 2.1.x is that 2.1.x is "development"
1150377624 M * harry and you don't need grsec on development ;)
1150378343 J * gerrit ~gerrit@67.160.146.170
1150379877 Q * Zaki Remote host closed the connection
1150380950 M * doener hm, the network namespace patches on lkml seem to have less overhead for "lo" than the delta-lo patches (I'm trying to be smart with that guess ;)
1150380968 M * doener still, the guy from IBM said that that patch produces a 1.5% overhead for lo
1150381017 M * doener "lo" being 127.0.0.1 in his case I assume...
1150381029 M * doener s/his/this/
1150381063 Q * gerrit Remote host closed the connection
1150381107 M * doener OTOH, the patches on lkml have probably a even worse overhead for anything else, because in quite a few places, you loose shortcuts for anything but 127.0.0.1
1150381116 J * gerrit ~gerrit@67.160.146.170
1150381127 M * doener the delta-lo patches seem to have the same overhead regardless of destination
1150381582 M * doener btw: http://lkml.org/lkml/2006/6/9/349 -- that's the patch series I'm talking about...
1150381791 Q * Smutje Quit: leaving
1150382526 J * Hollow ~hollow@home.xnull.de
1150382544 M * Hollow hey folks ;)
1150382553 M * jhaig I have tried building an rpm of the kernel with vserver built in but when I try to install it I get "grubby fatal error: unable to find a suitable template".  Maybe this isn't strictly a vserver question, but is there anything other than "make rpm" I need to do to make an rpm?
1150383419 M * yang is it possible that someone with root access on vserver, could be abusing (using them on IRC) IPs also from another guest vserver? 
1150383485 M * yang I mean, if interfaces can be abused also from another guest
1150383558 M * harry biiiiiiiiiiiig q here!
1150383576 M * harry (very big vserver problem too ;))
1150383590 M * harry i mount all "virtual hosts" in /vserver/<hostname> at boot
1150383614 M * harry then, all vservers start, so all mounts are known in all contexts
1150383632 M * harry now, when i shutdown, it stops all vservers (i hope)
1150383655 M * harry it unmounts all filesystems... (BUT!!! also in those contexts??? i don't know what happens here...)
1150383671 M * harry but when it tries to deactivate lvm, it says all volumes are still in uwe
1150383693 M * harry mdadm fails too... and when i reboot, all filesystems are NOT cleanly unmounted and require a filesystem check
1150383699 M * harry why's that? and what can i do?
1150383742 M * harry it's really messing up my system...
1150383754 M * harry is there a way to unmount stuff?
1150383780 J * Viper0482 ~Viper0482@p5497759B.dip.t-dialin.net
1150383821 J * pisco ~pampel@80.135.174.144
1150384180 Q * insomniac Ping timeout: 480 seconds
1150384335 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at
1150384739 J * insomniac ~insomniac@slackware.it
1150384873 M * Hollow harry: mount these filesystem only when the vserver is started
1150384883 M * Hollow not on system boot
1150385260 M * doener Hollow: hm, to really solve it, the mount would also have to happen in the vserver's namespace...
1150385304 M * doener otherwise vserver 1 only has its lv mounted, vserver 2 has its lv and that of vserver 1, etc...
1150385344 J * mkhl ~mkhl@200.148.40.47
1150385364 M * harry Hollow: not an option
1150385376 M * Hollow doener: well, i assumed that the / entry in the vservers fstab will be mounted in the namespace
1150385417 M * Hollow but i never actually used it
1150385420 M * Hollow so.. *shrug*
1150385435 M * harry i want my base system to be able to backup all the data
1150385441 M * Hollow harry: well, than you have to take care that all vservers are shut down before unmounting any filesystems
1150385442 M * harry just backup /vservers
1150385453 M * harry if i mount them at each startup... that's impossible
1150385460 M * doener it probably is, but you didn't say that he should use the fstab, but just "mount it then" ;)
1150385475 Q * cdrx Ping timeout: 480 seconds
1150385478 M * Hollow doener: ok :)
1150385488 M * doener harry: it only affects the file systems mounted under /vserver/<...>
1150385489 M * doener ?
1150385536 M * harry yes
1150385539 M * harry well... no
1150385542 M * harry sry :)
1150385544 M * harry so: no
1150385578 M * harry all mounted filesystems that are there at boottime are passed on to the vserver instances
1150385598 M * doener I meant the unclean lvm stuff
1150385632 Q * ||Cobra|| Remote host closed the connection
1150385641 M * Hollow actually, i have the same setup, i never had any problems unmounting stuff
1150385673 M * Hollow but well, the machine only rebooted once since i installed it :o
1150385676 M * doener harry: let's get back to step a) ;)
1150385685 M * doener harry: util-vserver from source or some distro package?
1150385737 M * harry source off course ;0
1150385740 M * harry ;)
1150385753 M * harry i think i found the reason tough...
1150385758 M * doener ok, and all your vserver's are tagged for the init-script to recognize them?
1150385763 M * harry not all vservers start by default :)
1150385768 M * harry nope :)
1150385777 M * doener because vanilla util-vserver has no "shut down all running vservers"
1150385777 M * harry that's why it doesn't work properly i presume :)
1150385787 M * harry i need that tough :)
1150385861 M * doener http://paste.linux-vserver.org/90
1150385867 M * doener that's how debian does it
1150385924 M * Hollow ah, ic..
1150385937 M * Hollow gentoo init script does so too
1150385968 M * daniel_hozac yang: no.
1150386062 M * daniel_hozac i've been meaning to write a proper initscript too, i guess i don't have to :)
1150386141 M * daniel_hozac is start-vservers --stop --all broken as well? seems to work fine here...
1150386227 M * daniel_hozac harry: and distro packages of util-vserver will usually include waldi's namespace cleanup patch, so all you have to do is touch /etc/vservers/.defaults/namespace-cleanup to resolve the "all mounts are copied to the namespace" problem.
1150386245 M * daniel_hozac (which i found is a bit broken btw, umount vs. umount -n)
1150386312 M * Hollow daniel_hozac: not sure, but the --stop --all is/was broken in some way too
1150386331 M * waldi daniel_hozac: and -l
1150386342 M * daniel_hozac well, util-vserver-0.30.210-start-vservers.patch is meant to fix that ;)
1150386357 M * waldi it reports errors if /usr is a different filesystem
1150386368 M * daniel_hozac waldi: yeah, that makes sense too.
1150386395 M * waldi i hope that works as expected ...
1150386405 M * Hollow daniel_hozac: ah?
1150386422 A * Hollow goes looking
1150386452 P * pisco 
1150386521 M * Hollow hm.. not in gentoo, bad me
1150386721 M * daniel_hozac waldi: http://daniel.hozac.com/vserver/delta-namespace-cleanup.patch
1150386734 M * harry daniel_hozac: how do i know if that works???, the namespace_cleanup stuff
1150386735 M * harry ?
1150386761 M * daniel_hozac vnamespace -e <guest> cat /proc/mounts?
1150386784 M * harry i will probably have to restart the vservers then...
1150386799 M * daniel_hozac of course, it won't be doing cleanup on running guests.
1150386810 M * daniel_hozac shouldn't be hard to get the code from it and do it manually though.
1150386815 M * harry is there a way to check if i have it, in the code?
1150386820 M * harry my util-vserver code?
1150386833 M * harry (i dont want to needlesly reboot)
1150386853 M * daniel_hozac grep isNamespaceCleanup /usr/lib*/util-vserver/vserver.functions
1150386901 M * harry damn... not there
1150386919 M * harry i could make all vserver instances start by default too, right?
1150387012 M * harry can i just : echo default > /usr/local/etc/vservers/.defaults/apps/init/mark             
1150387018 M * harry ?
1150387030 M * daniel_hozac i don't think so.
1150387076 M * harry why not? why does that work with /etc/vservers/.defaults/namespace-cleanup ?
1150387081 M * daniel_hozac nope, doesn't look like it.
1150387088 M * daniel_hozac because .defaults is special cased for every option.
1150387160 M * harry ah
1150387191 M * waldi daniel_hozac: looks okay, but does the files already use bash explicitely?
1150387200 M * harry so... with this http://daniel.hozac.com/vserver/delta-namespace-cleanup.patch patch, applied to util-vserver 0.30.210, i won't have that unmount problem anymore, right?
1150387484 M * daniel_hozac waldi: yeah, pretty much everything else is an array.
1150387484 Q * insomniac Read error: Connection reset by peer
1150387502 M * daniel_hozac harry: no, that's against the old namespace-cleanup patch.
1150387509 M * harry ah
1150387513 M * daniel_hozac http://daniel.hozac.com/vserver/util-vserver/util-vserver-0.30.210-namespace-cleanup.patch is the actual patch.
1150387516 M * harry so ... what do i do, to solve my prob
1150387520 M * harry lemme seee :)
1150387540 M * harry what are all the other patches???
1150387558 M * harry i'm now using 0.30.210
1150387561 M * harry (vanilla ;))
1150387562 M * daniel_hozac various fixes.
1150387598 M * harry hmm... what stops you from applying all of them, and calling it 0.30.211 ?
1150387618 M * daniel_hozac the fact i'm still considering Enrico upstream? :)
1150387626 M * harry ?
1150387647 J * insomniac ~insomniac@slackware.it
1150387686 M * harry are there other patches i should apply to get things running... well... better? :)
1150387812 M * daniel_hozac only if you have run into the issues they fix :)
1150387848 M * harry like the chattr stuff?
1150387897 M * harry http://daniel.hozac.com/vserver/util-vserver/util-vserver-0.30.210-start-vservers.patch
1150387901 M * harry what's with that one???
1150387907 M * harry do i need that too?
1150387914 M * harry (/me scanning all the patches now ;))
1150387920 M * daniel_hozac it makes --all really mean all.
1150387962 M * daniel_hozac and makes sure it removes all files it creates.
1150387995 M * harry so i should use that one too?
1150388007 M * harry and the condrestart ?
1150388015 M * daniel_hozac if you use start-vservers in a non-default way.
1150388049 M * harry mkay... now one that interests me
1150388053 M * harry the prefix patch?
1150388058 M * harry what does that one do?
1150388075 M * daniel_hozac yum breaks if you ./configure with --prefix=/ rather than --prefix=
1150388100 M * harry hmm... its got nothing to do when you compile from source then...
1150388135 M * daniel_hozac uh, it's only relevant if you compile from source.
1150388145 M * harry i do compile from source
1150388147 M * daniel_hozac AFAIK there's no distro configuring it with --prefix=/ or --prefix=...
1150388165 M * harry i configure with --prefix=/usr/local
1150388210 M * harry how about vlogin.c
1150388212 M * harry ?
1150388268 J * dhansen ~dave@sprucegoose.sr71.net
1150388270 J * bonbons ~bonbons@83.222.39.166
1150388335 M * daniel_hozac vlogin allocates a new psuedo tty inside the guest, rather than using the host's.
1150388348 M * harry why would i need that? ;)
1150388361 M * harry (/me doing millions of things at once now... sry ;))
1150388511 M * daniel_hozac well, vserver ... enter can run screen for instance.
1150388621 Q * jhaig Quit: Download Gaim: http://gaim.sourceforge.net/
1150388798 M * daniel_hozac Hollow, doener: start-vservers --running --stop seems to work here as well... probably due to the same fix.
1150389076 M * doener were any of your patches send to ensc/accepted?
1150389121 M * daniel_hozac quite a few of them are in reports on savannah
1150389138 M * daniel_hozac ensc hasn't responded to any of them though.
1150389309 M * daniel_hozac i've been meaning to send him an email.
1150389459 M * harry messed up my system... :S
1150389471 M * harry the  util-vserver-0.30.210-namespace-cleanup.patch  
1150389814 M * daniel_hozac how so?
1150389833 M * daniel_hozac it seems to be working fine here.
1150390504 J * alaa ~alaaomar@81.10.117.250
1150390897 M * harry well... since it's a production machine, i can't really test and elaborate on it...
1150390904 M * harry but no vserver wanted to start
1150390938 Q * alaa Quit: Leaving
1150391128 J * restill ~restill@c-71-197-23-172.hsd1.mi.comcast.net
1150391141 M * restill Hey all.  Long time
1150391624 M * daniel_hozac hey.
1150391671 P * meandtheshel1 
1150391691 M * restill Hey daniel.  Do you run the mailing list?
1150391753 M * daniel_hozac no, that's Martin List-Petersen, IIRC.
1150391785 M * daniel_hozac why?
1150391918 M * restill My boss wants me to setup a Usenet type thing for something he is working on.  I was going to try to use nntp and mailman
1150392348 J * meandtheshel1 ~markus@85-124-174-184.dynamic.xdsl-line.inode.at
1150392367 Q * genosid Quit: 
1150392394 Q * shedi Quit: Leaving
1150393563 M * restill have a good day.  ttyl
1150393565 P * restill Leaving
1150394396 Q * yang Ping timeout: 480 seconds
1150394832 N * sars sarnold
1150395259 Q * mkhl Quit: 
1150395330 Q * independence Ping timeout: 480 seconds
1150395585 Q * ddlp Ping timeout: 480 seconds
1150396153 J * s0undt3c1 ~s0undt3ch@bl8-4-240.dsl.telepac.pt
1150396212 Q * doener Quit: leaving
1150396522 J * prae ~benjamin@sherpadown.net
1150396595 Q * s0undt3ch Ping timeout: 480 seconds
1150397250 J * EmRe_17 huseyin@88.233.15.80
1150397741 J * glen_ ~glen@elves.delfi.ee
1150397787 M * glen_ hello. what is the reason, that using external package management on amd64 the vrpm says that it cant find groups, while the etc/groups does contain the id
1150397801 N * glen_ glen
1150397846 M * glen i'm using 0.30.210 tools version
1150398511 J * gerrit_ ~gerrit@67.160.146.170
1150399056 Q * dna Quit: Verlassend
1150399154 Q * derjohn2 Ping timeout: 480 seconds
1150399160 J * derjohn2 ~aj@dslb-084-058-213-087.pools.arcor-ip.net
1150399303 J * dna ~naucki@dialer-160-146.kielnet.net
1150399617 J * pisc1 ~pampel@80.135.162.234
1150400171 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1150400509 M * daniel_hozac glen: hmm, example?
1150400564 J * doener ~doener@i5387C3CD.versanet.de
1150400824 M * glen daniel_hozac: rpm installs package who has in %files %attr(600,root,groupname), and then rpm says can't find groupname, using root instead
1150400862 M * daniel_hozac i get that, but i meant an example pacakge.
1150400896 M * daniel_hozac one which exposes the issue.
1150400910 M * daniel_hozac is your guest a 64 or 32-bit install? of what distro?
1150400936 M * glen both are 64bit
1150400939 M * glen pld linux
1150400975 M * daniel_hozac ah.
1150401059 M * daniel_hozac and this group exists in the guest?
1150401067 M * daniel_hozac or is it supposed to be created by the %pre script?
1150401154 M * matled_ is there any problem mounting a filesystem that was in use with a vserver patched kernel with an unpatched kernel later?
1150401178 N * matled_ matled
1150401185 M * glen daniel_hozac: yes. exactly
1150401208 M * glen daniel_hozac: that means it's created in  %pre, and now it exists in guest
1150401411 M * glen and btw, i created same gid in host too, it didn't help
1150401417 M * doener matled: AFAICT there are no problems... I switch between vanilla and vserver kernels rather often... those mounted with tagxid might have strange uids/guids though
1150401536 P * meandtheshel1 
1150401573 M * matled I haven't done anything with vservers on this computer yet, but the last kernel I used just complained that there were unsupported optional features on the filesystem, dunno what this was..
1150401596 M * matled so without tagxid the filesystem should not get anything special with the vserver patch?
1150401935 M * doener right
1150402313 Q * fwl Quit: This computer has gone to sleep
1150402633 J * DreamerC_ ~dreamerc@59.112.8.20
1150403026 Q * DreamerC Ping timeout: 480 seconds
1150403320 J * f_ ~f_@83-215-237-1.seek.stat.salzburg-online.at
1150403535 Q * EmRe_17 Quit: Leaving
1150404106 Q * bonbons Quit: Leaving
1150404290 Q * zkbrsnie Quit: 
1150405604 Q * lilalinux Ping timeout: 480 seconds
1150405813 J * shedi ~siggi@130.208.221.254
1150405919 M * daniel_hozac glen: i can't reproduce that here. are you doing anything special?
1150406109 J * lilalinux ~plasma@dslb-084-058-219-071.pools.arcor-ip.net
1150406230 M * daniel_hozac glen: but that's on Fedora.
1150406522 Q * f_ Quit: This computer has gone to sleep
1150406534 Q * dna Quit: Verlassend
1150406565 Q * Viper0482 Remote host closed the connection
1150407733 Q * prae Quit: Pwet
1150408095 J * Aiken ~james@tooax7-054.dialup.optusnet.com.au
1150408812 J * Zaki ~Zaki@88.213.19.115
1150408881 M * glen daniel_hozac: i don't think so. there are some patches, but none of them is causing imho this
1150408894 M * glen altho i'm using poldek to install, maybe vpoldek isn't integrated properly
1150408980 M * daniel_hozac that's not a vanilla thing, so i really can't help you with that.
1150409036 M * glen let me see if vrpm works alone okay
1150409124 M * glen ok. fails too
1150409133 Q * lilo2 Quit: bbiab
1150409156 M * glen # vrpm cvs -- -Uhv masqmail-0.2.20-5.amd64.rpm --force
1150409157 M * glen Preparing...                ########################################### [100%] 1:masqmail               ########################################### [100%]
1150409160 M * glen warning: user mail does not exist - using root
1150409164 M * glen # grep mail etc/group
1150409164 M * glen mail::12:mail
1150409164 M * glen uucp::14:uucp,mail
1150409190 M * glen ah sorry, wrong grep, here's correct:
1150409191 M * glen # grep mail etc/passwd
1150409191 M * glen mail:x:8:12:mail:/var/mail:/bin/false
1150409336 M * glen maybe it's rpm bug
1150409348 M * glen or glibc
1150409348 M * daniel_hozac can you reproduce it with vanilla util-vserver?
1150409441 M * glen i've ran it with with some debug
1150409442 M * glen http://glen.alkohol.ee/pld/useradd-vserver.txt
1150409553 M * glen ok. perhaps this brokes it? http://cvs.pld-linux.org/cgi-bin/cvsweb.cgi/SOURCES/util-vserver-vrpm.patch
1150409592 M * daniel_hozac that one actually looks sane.
1150409665 M * glen it doesn't happen on i686. at least i haven't noticed
1150409765 M * glen it's userspace tool problem right?
1150409803 M * daniel_hozac i'd say so.
1150409994 M * glen happens with vanilla too (i left out all patches)
1150410181 M * glen but it doesn't happen on other amd64 host
1150410267 Q * cdrx Quit: Leaving
1150410687 M * glen i traced it down to this single command: /usr/sbin/vnamespace --enter 2401 -- /usr/sbin/vnamespace --new -- /usr/lib64/util-vserver/vrpm-worker cvs -Uhv masqmail-0.2.20-5.amd64.rpm --force
1150410800 M * daniel_hozac the vnamespace --enter vnamespace --new looks really odd.
1150410839 M * glen but the same command was also executed on host where are no problems
1150410857 M * daniel_hozac so if you execute that command manually, you don't get any problems?
1150410882 M * glen no. nothing has changed. works on one host and doesn't other
1150411060 M * daniel_hozac hmm. it seems rpm-fake isn't used with vrpm...
1150411062 M * glen the kernel is different for them, userspace tools in host are same
1150411073 M * daniel_hozac (or i just suck and i can't find it)
1150411143 M * glen ++ _RPM_FAKE_SO=/usr/lib64/util-vserver/rpm-fake.so
1150411162 M * glen + exec /usr/lib64/util-vserver/vrpm-preload --root /vservers/cvs -Uhv masqmail-0.2.20-5.amd64.rpm --force
1150411167 M * glen so perhaps it's in vrpm-preload
1150411193 M * daniel_hozac ah, so i do suck :)
1150411433 M * glen is there way to debug vrpm or rpm-fake?
1150411610 M * glen     uint32_t            id;
1150411619 M * glen is it 32bit on amd64? 
1150411632 M * daniel_hozac yes.
1150411707 M * daniel_hozac RPM_FAKE_CHROOT=/vservers/cvs RPM_FAKE_CTX=<xid> RPM_FAKE_CAP=$[ ~0xd40c04ff ] RPM_FAKE_FLAGS=4 LD_PRELOAD=/usr/lib64/util-vserver/rpm-fake.so rpm --root /vservers/cvs -Uhv masqmail-0.2.20-5.amd64.rpm?
1150411783 M * glen um, rpmdb is not present
1150411820 M * glen ok. i put --nodeps, and nothing. still can't get uid/gid
1150411824 M * daniel_hozac i guess you'll have to run vrpm-preload with debugging then to get the proper command.
1150411894 M * glen http://pastebin.com/711804
1150411923 M * glen i can't strace such proc...
1150411943 M * daniel_hozac indeed.
1150412011 M * glen compile with  --enable-lib-debug ?
1150412027 M * daniel_hozac how about LD_PRELOAD=/usr/lib64/util-vserver/rpm-fake.so RPM_FAKE_CHROOT=/vservers/<name> RPM_FAKE_CTX=<xid> RPM_FAKE_CAP=0 RPM_FAKE_FLAGS=0 id <user only present in the guest>
1150412048 M * daniel_hozac i doubt that helps.
1150412072 M * glen uh, the #ifdef i saw in rpm-fake.c is already #define DEBUG 1 in same file ;)
1150412110 M * glen failure
1150412134 M * daniel_hozac failure? as in doesn't find the user?
1150412140 M * glen actually i checked for uid which is present in both
1150412163 M * glen http://pastebin.com/711809
1150412194 M * daniel_hozac how very odd.
1150412223 M * glen indeed indeed
1150412241 M * glen hmm, 'id root' fails, but just 'id' works
1150412263 M * daniel_hozac well, id root queries nss.
1150412271 M * daniel_hozac does id work inside the guest?
1150412290 M * glen yes
1150412303 M * glen # chroot /vservers/cvs/ id mail
1150412303 M * glen uid=8(mail) gid=12(mail) groups=12(mail),14(uucp)
1150412320 M * glen # vserver cvs enter
1150412320 M * glen [root@pldmachine /]# id mail
1150412320 M * glen uid=8(mail) gid=12(mail) groups=12(mail),14(uucp)
1150412404 M * daniel_hozac so it's definitely rpm-fake breaking it.
1150412426 M * daniel_hozac could you try updating to the latest syscall implementation?
1150412445 M * daniel_hozac i had issues with rpm-fake being miscompiled before.
1150412453 M * glen that means what? userspace or kernel?
1150412467 M * daniel_hozac http://daniel.hozac.com/vserver/util-vserver/util-vserver-0.30.210-shiny10.patch
1150412823 M * glen didn't help
1150412874 Q * DreamerC_ Quit: leaving
1150412903 M * glen could i test is rpm-fake-resolver working?
1150412912 J * DreamerC ~dreamerc@59.112.8.20
1150413013 M * glen i see it makes some socket, it's for communicating with rpm-fake-resolver?
1150413031 M * daniel_hozac yeah.
1150413174 M * glen # /usr/lib64/util-vserver/rpm-fake-resolver -r /vservers/cvs/ -c 2401
1150413175 M * glen rpm-fake-resolver: write(): Bad file number
1150413182 M * glen but i'm probably just invoking it wrong
1150413235 M * daniel_hozac you probably need to setup some fds.
1150413836 M * glen # echo -n 'Gmail' | /usr/lib64/util-vserver/rpm-fake-resolver -r /vservers/cvs/ -c 2401 3>&1
1150413839 M * glen ..
1150414074 M * glen so fake-resolver doesn't work?
1150414131 M * daniel_hozac are you sure you're setting up the fds correctly?
1150414160 M * glen yes. as i sent '.' it responded with '.' and 'Q' exited immediately
1150414182 M * glen at least so i understhood how run() in ../BUILD/util-vserver-0.30.210/src/rpm-fake-resolver.c works
1150414253 M * daniel_hozac doesn't that mean it works?
1150414308 M * daniel_hozac i guess you could hack rpm-fake-resolver to do some logging with debugging info.
1150414324 M * glen right. it calls readall, so it probaly read all the input
1150414369 M * glen altho Ereadall has parame sizeof(c), which could be 1 byte in case of uint8_t
1150414393 M * glen just write to stderr?
1150414402 M * daniel_hozac readall just means it will _really_ read that much data.
1150414418 M * daniel_hozac (i.e. if there's a short read, it will read again, starting at where the previous read stopped)
1150414735 M * glen i tried the same call on host where rpm worked, and same failure, so i most likely tested wrong
1150415081 M * glen ahaa, G + length + data!
1150415124 M * daniel_hozac yep.
1150415343 M * glen echo -ne 'G\004mail' should had been working, but didn't
1150415614 M * glen right, int4
1150415673 M * glen getgrnam returned 0
1150415784 M * glen http://pastebin.com/711911
1150415813 M * glen that ÿ is chr(255) aka -1
1150415913 M * glen getgrnam() says in strerror(): Invalid argument
1150415921 M * glen ideas?
1150415952 M * daniel_hozac invalid argument from getgrnam? that's bizarre.
1150415956 J * doener_ ~doener@i5387CB47.versanet.de
1150415957 M * daniel_hozac you are using dietlibc, right?
1150415970 M * glen yes, diet
1150415990 M * glen i'll try now with glibc