1161043220 M * Bertl rgl: can't really tell, it's kind of gut feeling, I do not feel comfortable with hg
1161043245 M * Bertl besides that, it is absolutely unsuited for kernel development
1161043249 M * Bertl *development
1161043270 M * rgl it is?
1161043275 A * Bertl .o( hmm, I start correcting correct spelling :)
1161043283 M * rgl what are you missing in it?
1161043304 M * Bertl well, for kernel development, you need to have thousands of kernel branches at your fingertips
1161043330 M * Bertl you need to be able to switch between them very easily, and they must not fill up your disk
1161043350 M * Bertl (for several reasons, the disk space is really important ... think caches and such)
1161043364 M * rgl that seems like a hg match, I don't understand :|
1161043419 M * Bertl okay, how would I check out a kernel tree from hg, without creating a 'new' copy of that tree
1161043441 M * Bertl i.e. just something I would get with cp -la (hard linked with other kernel trees)
1161043478 M * rgl humm, I don't think that checkouts are hardlinked.  the repo itself is
1161043488 M * Bertl see, that's the problem
1161043507 M * Bertl when I operate on a 'copy' everything gets slow
1161043580 M * rgl and git can do that?
1161043589 M * Bertl yep, git is basically all hard links
1161043615 M * rgl even the working copy?  thats odd
1161043638 M * rgl what happens when you change a file?
1161043649 M * rgl all the other links are changed too?
1161043667 M * Bertl nope, proper editor configuration takes care of that
1161043727 M * rgl humm you never killed a file by accident? hehe
1161043765 M * Bertl it happened, but only once or twice over the last few years
1161043857 M * rgl not bad :)
1161043871 M * Bertl but I just saw there was a patch for hg which can do that too, if you are interested ...
1161043894 M * rgl if you got a link at hand :)
1161043900 M * Bertl http://www.selenic.com/pipermail/mercurial/2005-June/001468.html
1161043931 M * Bertl you see the advantage in the timing :)
1161043967 M * rgl and git does that by default, or you had to change/give a specific option?
1161043985 Q * derjohn2 Remote host closed the connection
1161044107 M * hardwire wazzuuuuuu
1161044109 M * hardwire :)
1161044122 M * Bertl rgl: git is all options and it is still in flux
1161044307 M * rgl Bertl, ah ok.
1161044317 M * rgl I gtg.  its late here :(
1161044323 M * rgl g'night
1161044329 M * Bertl have a good one!
1161044410 J * litage ~nick@203.220.55.70
1161044419 M * Bertl welcome litage!
1161044441 M * litage howdy Bertl. it's been a while  ;)
1161044451 Q * rgl Read error: Connection reset by peer
1161044558 M * Bertl litage: yep, indeed ... how's going?
1161045910 M * cehteh offtopic, but can someone give me a hand with a gigabit ethernet problem?
1161046084 M * Bertl hmm?
1161046174 M * litage Bertl: sorry, got a million things going on  :P  things are going pretty well. most importantly, the vservers are still up and kicking!
1161046208 M * Bertl litage: good, that's how it's supposed to be! :)
1161046223 M * litage haha yeah exactly
1161046237 M * litage and i'm running a version that's fairly old, too
1161046256 A * cehteh is getting gbit connection in one direction but only fast ethernet the way back
1161046257 M * litage when i restart syslog-ng in a vserver guest, this error occurs:  "/etc/cron.daily/logrotate: klogctl: Operation not permitted"    in non-vserver boxes, setting CONSOLE_LOG_LEVEL=1 in /etc/default/syslog-ng fixes this, but it doesn't seem to in vserver guests
1161046299 M * cehteh litage: just a hint, you can do centralized sysloging from the root server
1161046300 M * Bertl cehteh: can happen, if one pair is faulty
1161046330 M * cehteh Bertl: any idea how to check which one? 
1161046332 M * Bertl gbit uses all 4 pairs for transmission two in each direction
1161046340 M * cehteh yes i know ..
1161046344 M * Bertl (given that you are talking about copper :)
1161046361 M * cehteh but autonegotation shows gbit .. no direction mentioned
1161046365 M * cehteh yep
1161046386 M * Bertl what about ethtool output on both sides?
1161046390 M * cehteh well .. i think i just exchanging cables
1161046398 M * cehteh theres a switch between
1161046417 M * Bertl which shows gbit on both plugs?
1161046418 M * cehteh and ethtool reports 1000mbit for both computers
1161046424 M * cehteh yes that too
1161046440 M * Bertl that's at least funny, my switch figured a bad cable very fast
1161046451 M * cehteh actually all 4 plugs .. computer and switch
1161046499 M * Bertl try with a crossover cable first (or normal patch cable if auto crossing on at least one card)
1161046519 M * Bertl to rule out misconfigurations and/or card issues
1161046554 M * cehteh mhm .. that involved much cableing and running between the desks ..currently working remote on both computers from my laptop :P
1161046651 M * cehteh cable exchaning is simple ... leme try
1161046907 M * cehteh no change
1161046961 M * Bertl how do you 'measure' the throughput, btw, if you say, both cards show gbit?
1161046966 M * cehteh iperf
1161046996 M * Bertl what does the cpu usage show on both machines when you do that?
1161047064 M * cehteh didnt checked .. but cant be the problem, both are fast machines amd64 with chipset gbit connected to the hypertransport
1161047098 M * Bertl okay, what about the switch ports?
1161047110 M * cehteh well i suspect flaky drivers .. nforce2 got quite some overhaul recently and i just wait until 2.6.19 is released
1161047131 M * Bertl (i.e. some ports can be configured to do fast ethernet only, could also be a faulty port)
1161047149 M * cehteh and the other one is a Realdreck ... ethtool shows that most things are not supported and the kernel is little older
1161047172 M * Bertl hehe
1161047186 M * cehteh yeah but its all acknowledged as 1000mbit
1161047216 M * cehteh and i never heard of ports which can be async configured to send 100mbit but recive 1000 or vice versa
1161047244 M * Bertl you could get this 'rare' setup with a monitoring port though :)
1161047291 M * cehteh hey my new switch is plug'n'play! ... (marketing jargon for unmanaged)
1161047312 M * Bertl web interface?
1161047319 M * cehteh unmanaged!!
1161047334 M * Bertl really? they do still build unmanaged gbit switches?
1161047336 M * cehteh plugnplay .. nothing else
1161047341 M * cehteh yes ;)
1161047347 M * Bertl fascinating!
1161047356 M * cehteh well would be ok ... for soho 
1161047378 M * cehteh and i really dont suspect the switch is the problem
1161047391 M * cehteh tomorrow i try a direct link
1161047401 M * Bertl okay, well, the drivers leave a lot of room of issues :)
1161047415 M * Bertl both nforce and realtek are somewhat special
1161047420 M * cehteh hehe
1161047422 M * cehteh ack
1161047437 M * Bertl I usually go for intel, (very old digital and 3com), or marvel
1161047452 M * cehteh oh man .. i love my 2 macronix cards .. cheap as realtek, fast and reliable as eepro
1161047474 M * cehteh unfortunally they are rare
1161047483 M * cehteh and not gigabit
1161047510 M * cehteh but ack the old eepro100 where the best ones ever made
1161047520 M * cehteh for fast ethernet
1161047532 M * Bertl the digital cards were the best ones out ther IMHO
1161047569 M * cehteh digital doesnt count ... they are too rare
1161047586 M * Bertl well, for some time, you could get them easier and cheaper than the 3com
1161047595 M * cehteh really uhm
1161047597 M * Bertl granted, it was a short period of time
1161047602 M * cehteh heh
1161047625 M * cehteh ad my both macronix cards
1161047677 M * cehteh realtek screwed them all with quantity over quality
1161047701 M * Bertl well, the realtek design was _two_ steps in the wrong direction :)
1161047719 M * Bertl but they are still selling cards :)
1161047728 M * cehteh they are cheap 
1161047758 M * cehteh means there is barely any silicon on them .. most work does the cpu
1161047791 M * Bertl that would be fine, but the buffer design is flawed in so many ways ...
1161047807 M * Bertl but hey, design flaws are the future :)
1161047830 M * cehteh heh
1161047850 M * Bertl recently I noticed that the (newer) nvidia cards do not have any color/brightness/contrast controls for video overlay ... that's supposed to happen in the cpu now
1161047891 M * Bertl I'd grant an extra price to the person who figured that (for stupidity)
1161048193 M * cehteh eww
1161050708 J * Aiken_ ~james@tooax6-218.dialup.optusnet.com.au
1161051037 Q * Aiken Ping timeout: 480 seconds
1161053706 Q * ensc Killed (NickServ (GHOST command used by ensc_))
1161053716 J * ensc ~irc-ensc@p54B4F7EB.dip.t-dialin.net
1161056342 Q * bronson Quit: Ex-Chat
1161056663 Q * rob-84x^ Ping timeout: 480 seconds
1161057073 Q * Aiken_ Quit: Leaving
1161057144 J * Aiken ~james@tooax6-218.dialup.optusnet.com.au
1161057871 M * Bertl daniel_hozac: are there any known vlogin issues with 0.30.211?
1161058008 M * Bertl on powerpc I observe something strange ... I can enter quite fine, but when I type stuff, nothing is written/executed unless I press CTRL-D at which point the typed commands are executed quite fine (note: this is on a 2.0.2.2 prerelease)
1161058882 Q * glut Read error: Operation timed out
1161059147 M * Bertl daniel_hozac: interesting detail, it doesn't happen with a chcontext style 'enter' and everything works fine with ssh logon (as expected), will upload a release candidate now (for testing)
1161059205 M * daniel_hozac Bertl: hmm, another arch with broken cfmakeraw maybe?
1161059221 M * Bertl could be, how would I check/know?
1161059256 M * daniel_hozac well, the sparc issues were mostly found by accident, when just checking the dietlibc headers against the kernel headers.
1161059634 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.18.1-vs2.0.2.2-rc1.diff
1161059645 M * Bertl okay, the following 'options' in that rc1
1161059679 M * Bertl in kernel/pid.c (pid_task()) are two commented out alternatives
1161059700 M * Bertl changing both of them makes it _really_ restrictive
1161059750 M * Bertl all variants seem to work quite fine in regard of init/fake and 'normal' signalling/ssh logon/chcontext
1161059799 M * Bertl daniel_hozac: if you find some time, I'd appreciate a test run on x86 or x86_64 with your tools, just to get an idea if the vlogin works there ...
1161059844 M * Bertl btw, I decided to add the enter/leave stuff, but I haven't verified that I didn't miss something, so a double check would be appreciated too
1161060265 M * daniel_hozac hmm, should the warnings for pid_task really be there?
1161060281 M * Bertl what does it warn about?
1161060294 M * daniel_hozac WARN_ON(current->xid) when extrapolated.
1161060303 M * daniel_hozac seems like the kernel is going to be very verbose :)
1161060329 M * Bertl well, no, it should not really happen, but yeah, it's for debugging only (atm)
1161060355 M * Bertl the idea is to move the restrictive part into the warn condition
1161060382 M * Bertl i.e. block early (with result = NULL) but only warn on unusual conditions
1161060383 M * daniel_hozac hmm, pid_task won't be called from guests?
1161060401 M * Bertl it will, but the condition is rare
1161060413 M * Bertl (only when you have cross context lookups)
1161060420 M * daniel_hozac ah
1161060449 J * bronson ~bronson@c-71-198-75-160.hsd1.ca.comcast.net
1161060464 M * daniel_hozac i'll give it a spin when i get back from school today, see if i get any vlogin issues.
1161060468 M * Bertl one allowed constellation atm is the fake init blend-through
1161060482 M * Bertl excellent, tx! I'm off to bed now ...
1161060487 M * daniel_hozac ok, good night!
1161060497 M * Bertl tx, cya!
1161060505 N * Bertl Bertl_zZ
1161063578 J * rob-84x^ rob@submarine.ath.cx
1161066008 J * meandtheshell ~markus@85-125-230-183.dynamic.xdsl-line.inode.at
1161068092 Q * Aiken Ping timeout: 480 seconds
1161068878 J * dna_ ~naucki@156-196-dsl.kielnet.net
1161069828 Q * bronson Quit: Ex-Chat
1161070821 Q * m4z Read error: Connection reset by peer
1161070830 J * m4z m4z@bastard-operator.from-hell.net
1161071184 Q * ntrs Read error: Operation timed out
1161071502 J * glut glut@no.suid.pl
1161072175 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com
1161074030 Q * shedi Quit: Leaving
1161075040 J * aloneMalibog ~ander@tkt34.chikka.com
1161076298 J * Rusty` ~rusty@a1052.adsl.pool.eol.hu
1161076300 M * Rusty` hello guys
1161076314 M * Rusty` is here anybody? :)
1161076320 M * phedny always ;)
1161076468 M * Rusty` i have a big problem :(
1161076470 M * Rusty` apollo:~# vserver mailwall start
1161076470 M * Rusty` mount: none already mounted or . busy
1161076470 M * Rusty` /etc/vservers/mailwall/fstab:1:1: failed to mount fstab-entry
1161076470 M * Rusty` mount: none already mounted or . busy
1161076470 M * Rusty` /etc/vservers/mailwall/fstab:2:1: failed to mount fstab-entry
1161076471 M * Rusty` chbind: vc_set_ipv4root(): Function not implemented
1161076471 M * Rusty` An error occured while executing the vserver startup sequence; when
1161076473 M * Rusty` there are no other messages, it is very likely that the init-script
1161076473 M * Rusty` (/etc/rc.d/rc 2) failed.
1161076475 M * Rusty` Common causes are:
1161076475 M * Rusty` * /etc/rc.d/rc on Fedora Core 1 and RH9 fails always; the 'apt-rpm' build
1161076477 M * Rusty`   method knows how to deal with this, but on existing installations,
1161076477 M * Rusty`   appending 'true' to this file will help.
1161076479 M * Rusty` Failed to start vserver 'mailwall'
1161076491 M * Rusty` the host system is debian, and i want run CentOS in vserver environment
1161076527 M * doener kernel and tools version?
1161076535 M * doener does testme.sh show any problems?
1161076635 M * Rusty` linux kernel 2.6.17.8, vs2.0.2-grsec2.1.9, util-vserver 0.30.211
1161076668 M * doener is CONFIG_VSERVER_LEGACY set?
1161076704 M * Rusty` yes
1161076707 M * Rusty` apollo:~# ./testme.sh
1161076707 M * Rusty` Linux-VServer Test [V0.16] Copyright (C) 2003-2006 H.Poetzl
1161076707 M * Rusty` chcontext: vc_new_s_context(): Function not implemented
1161076707 M * Rusty` chcontext failed!
1161076707 M * Rusty` invalid option `nid'.
1161076707 M * Rusty` Try 'chbind --help" for more information.
1161076709 M * Rusty` invalid option `nid'.
1161076709 M * Rusty` Try 'chbind --help" for more information.
1161076711 M * Rusty` chbind failed!
1161076711 M * Rusty` Linux 2.6.8-2-386 #1 Thu May 19 17:40:50 JST 2005 i686
1161076713 M * Rusty` Ea 0.30.204 273/glibc (DS*) <compat,v11,v13,fscompat,net,oldproc,olduts>
1161076713 M * Rusty` VCI: <none> (Tn)
1161076714 M * doener argh!
1161076715 M * Rusty` ---
1161076717 M * Rusty` ooooo
1161076720 M * Rusty` aaaa
1161076722 M * Rusty` i am lame
1161076729 M * Rusty` :(
1161076735 M * Rusty` bad kernel... :)
1161076745 M * Rusty` sorry
1161076745 M * doener do you still have any old style configs?
1161076755 M * doener ie. .conf + .sh file?
1161076758 M * Rusty` i'm booting an old kernel..
1161076771 M * doener oh, wow
1161076775 M * doener didn't even notice that
1161076782 M * doener yeah, that one won't work ;)
1161076789 M * Rusty` :)
1161076805 M * doener anyway, if you don't have any legacy config vservers, you can completely disable legacy 
1161076851 M * doener ie. enable legacy == no, and enable networking api (or whatever that option was called) == yes
1161076876 M * doener hm, and old tools are around as well
1161076879 M * doener 11:18:33 <Rusty`> Ea 0.30.204 273/glibc (DS*) 
1161076891 M * doener probably you forgot to remove the debian package
1161076959 M * Rusty` oh, debian package is old?
1161077011 M * doener the one in sarge is ancient
1161077054 M * doener backports.org has recent util-vserver packages IIRC, not sure if they have .211 yet
1161077294 J * rgl Rui@217.129.151.190
1161077298 M * rgl hello
1161077338 M * doener wb rgl 
1161077402 M * rgl are there rpm packages for vserver?
1161077494 M * doener daniel_hozac should have some for fc
1161077546 M * doener http://rpm.hozac.com/dhozac/
1161077591 J * shedi ~siggi@dsl-149-109-85.hive.is
1161077965 M * phedny a hint for anyone who wants to transfer a vps to another system with rsync
1161077969 M * phedny please use --numeric-ids :)
1161078069 M * rgl doener, cool thx :D
1161078316 J * xtomece1 ~xtomece1@nymfe03.fi.muni.cz
1161078348 J * _are_ ~are@62.112.159.81
1161078352 M * _are_ Hi
1161078432 M * xtomece1 hi, is it possible to install some application on one vserver and make it easily visible to all vservers? I found 'vhashify', but I don't know how to use it
1161078461 M * _are_ no, this won't work
1161078488 M * _are_ hasify does hardlinks for files that are on both (several, all) vservers
1161078499 M * _are_ it is emant to safe space
1161078549 M * xtomece1 I have to install it on all vservers and than hashify it?
1161078580 M * _are_ if it is just files at specific places and you know these places, you could e.g. use rsync. Only if it is not package-management handled software, though.
1161078591 M * _are_ afterwards yolu can indeed hasify it
1161078638 M * xtomece1 Thanx. Another question, how to use copy on write, please?
1161078663 M * doener use a recent version of vs2.1.1 and hashify, done ;)
1161078664 M * _are_ sorry, no idea, never used it.
1161078686 M * doener hashifying without CoW can be quite disruptive
1161078818 J * debugger Rui@217.129.151.190
1161078852 M * xtomece1 OK, I am testing version of vs2.1.1 and also older. What will it make if I hashify it and then reboot to non-CoW kernel?
1161078918 M * doener on vanilla the unified files will be simply immutable.
1161078946 M * doener on vserver kernel without CoW, they will be IUNLINK, ie. immutable but unlinkable (deletable)
1161078981 M * doener the later is sufficient for files managed by your package management system that are not normally modified (eg. libs, executables)
1161079013 M * doener but hashifying might affect other files as well, for which the non-CoW behaviour can cause troubles
1161079061 M * xtomece1 Thank you all very much
1161079068 M * doener you're welcome
1161079092 Q * xtomece1 Quit: Ukončuji
1161079235 Q * mcp Read error: Connection reset by peer
1161079251 J * mcp ~hightower@wolk-project.de
1161079272 Q * rgl Ping timeout: 480 seconds
1161079498 N * phedny phedny_
1161079793 P * aloneMalibog 
1161079854 M * eyck wow, acpi started working on 2.6.17+
1161079878 M * doener so we lost a 2.4-vserver user?
1161079894 M * doener s/a/the/ ? *SCNR*
1161080127 Q * MrX Ping timeout: 480 seconds
1161080175 J * phedny ~mark@phedny.vps.van-cuijk.nl
1161080269 Q * phedny 
1161080303 M * eyck not yet
1161080370 J * phedny ~mark@phedny.vps.van-cuijk.nl
1161080612 A * phedny wants to thank everybody who helped me, vserver machine now is up and running :)
1161081189 M * _are_ is there a sample /etc/vserver/.distribution and a sample command line to get up some fedora core 4 or 5 from within a debian host? I tried, but never having run a fc-any box native I am sort of lost and either get the guides ...
1161081189 M * _are_ on wiki wrong or they are way incomplete.
1161081238 M * daniel_hozac hmm?
1161081263 M * daniel_hozac FC6 and older should be supported OOTB with util-vserver 0.30.211.
1161081289 M * daniel_hozac FC4 and older with 0.30.210.
1161081338 M * _are_ then I am just to stupid, I guess. Is it possible to install fc-any directly from inet? and how would the vserver build line look like?
1161081351 M * _are_ first attempt has been to create a local repository
1161081367 M * daniel_hozac vserver fc5 build -m yum --context... -- -d fc5
1161081423 M * daniel_hozac Bertl_zZ: vlogin works fine on my x86 test box with 2.0.2.2-rc1.
1161081449 M * _are_ tried that, .210 tools, failed with fc4
1161081533 M * daniel_hozac failed how?
1161081556 M * _are_ let me get the error again...
1161083332 M * Rusty` oh, i have a little problem :(
1161083333 M * Rusty` apollo:~# vserver mailwall start
1161083333 M * Rusty` Can not find a vserver-setup at '/usr/local/etc/vservers/mailwall/'.
1161083333 M * Rusty` Possible solutions:
1161083333 M * Rusty` * fix the spelling of the 'mailwall' vserver name
1161083333 M * Rusty` * read 'vserver mailwall build --help' about ways to create a new vserver
1161083333 M * Rusty` * see 'vserver --help' for the syntax of this command
1161083360 M * daniel_hozac well, umm, does /usr/local/etc/vservers/mailwall exist?
1161083382 M * Rusty` oh.. no
1161083384 M * Rusty` i am lame
1161083385 M * Rusty` :)
1161083410 M * Rusty` it works now, sory
1161083776 N * debugger rgl
1161083795 M * rgl daniel_hozac, how can I use your rpm repository for centos4?
1161083810 M * rgl daniel_hozac, that is, http://rpm.hozac.com/dhozac/centos/4/
1161083812 M * daniel_hozac rgl: you shouldn't. it's seriously outdated.
1161083828 M * daniel_hozac what is it that you want RPMs of? util-vserver?
1161083841 M * rgl daniel_hozac, do you still have the src rpm for the kernel?
1161083846 M * daniel_hozac rpmbuild -tb util-vserver-0.30.211.tar.bz2 should give you that.
1161083869 M * daniel_hozac rgl: uh, for what kernel?
1161083894 M * daniel_hozac i never did kernels for CentOS.
1161083898 M * rgl daniel_hozac, I though you also had rpm for the kernel eheh
1161083928 M * rgl you known where can I get a rpm for the kernel?
1161083941 M * rgl the src rpm is good enough :D
1161083951 M * daniel_hozac for CentOS's kernel?
1161083960 M * daniel_hozac or do you mean with vserver included?
1161084053 J * yarihm ~yarihm@whitehead2.nine.ch
1161084202 M * _are_ daniel_hozac: took me a while, but have the error messages now. The Repository i generated is obviously broken, but I can well live wirth an inet rposiotory if that is possible and someone gives me a hiont how this is doen:
1161084204 M * _are_ Warning, could not load sqlite, falling back to pickle
1161084205 M * _are_ No Repositories Available to Set Up
1161084205 M * _are_ No Match for argument: glibc
1161084206 Q * matti Ping timeout: 480 seconds
1161084248 M * daniel_hozac what does yum --version say?
1161084258 M * daniel_hozac and what util-vserver version are you using?
1161084281 J * debugger Rui@217.129.151.190
1161084301 M * _are_ vserver 210
1161084313 M * _are_ #  yum --version
1161084313 M * _are_ Warning, could not load sqlite, falling back to pickle
1161084313 M * _are_ 2.4.0
1161084327 M * _are_ bare in mind, host is a debian system
1161084371 M * daniel_hozac could you try upgrading to 0.30.211?
1161084395 M * _are_ will running vservers be restarted or otherwise misbehave?
1161084442 M * daniel_hozac not unless you're using VLAN interfaces that you want to get removed/created on stop/start.
1161084474 M * _are_ no, i use eth aliases only
1161084667 Q * rgl Ping timeout: 480 seconds
1161084793 M * daniel_hozac _are_: and btw, that's all the output? you don't get anything about your yum version being bad?
1161085177 N * debugger rgl
1161085188 M * rgl daniel_hozac, with vserver included :)
1161085199 M * daniel_hozac rgl: AFAIK there are none.
1161085206 M * daniel_hozac CentOS's kernel is ancient.
1161085230 M * rgl it doesnt have to be centos kernel, it can be vanilla with vserver patch
1161085246 M * daniel_hozac i'm not aware of any such RPMs either.
1161085256 M * daniel_hozac although that would make sense.
1161085280 M * rgl what I really wanted is some src.rpm to bootstrap the thing, instead of creating one from scratch
1161085833 J * debugger Rui@217.129.151.190
1161085872 M * _are_ daniel_hozac: no message about yum being bad
1161086019 J * debugger_ Rui@217.129.151.190
1161086022 M * Borg- hmm weiriod
1161086030 M * Borg- [root@titanium] vserver titanium0 start
1161086030 M * Borg- vcontect: chroot(): Permission denied
1161086048 M * Borg- this happen when I have --barrier flag set on guest root dir
1161086057 M * Borg- when no barrier flag.. eveyrthing works
1161086069 M * daniel_hozac ... as expected.
1161086084 M * daniel_hozac you're supposed to set the barrier on the directory above the guest's root.
1161086191 Q * rgl Ping timeout: 480 seconds
1161086217 M * daniel_hozac _are_: that should be fixed in 0.30.211 then.
1161086237 M * Borg- daniel_hozac: ah.. ok :)
1161086244 J * matti matti@linux.gentoo.pl
1161086249 N * debugger_ rgl
1161086489 Q * debugger Ping timeout: 480 seconds
1161086764 Q * mire Ping timeout: 480 seconds
1161086930 M * daniel_hozac Bertl_zZ: http://people.linux-vserver.org/~dhozac/p/k/delta-headers-fix87.diff unneeded headers in 2.0.2.2-rc1
1161086977 J * mire ~mire@131-167-222-85.adsl.verat.net
1161089520 J * MrX ~urk@219.95.5.174
1161091227 J * gluk ~kvirc@breuss.ws.ehouse.ru
1161091679 Q * rgl Quit: Fui embora
1161092891 M * _are_ daniel_hozac: installed 211 tools, completly non-messed fc5 directory and it actually does something. actually it complains abouzt a messed up yum, then tzries downloading but fails on multiple servers. Gave http_proxy and ...
1161092891 M * _are_ ftp_proxy on the command line as there is a firewall, but seems this is not accounted.
1161093077 M * daniel_hozac so you have to use the proxy? http_proxy/ftp_proxy should work fine. do the files it's looking for exist?
1161093130 M * daniel_hozac i.e. can you get to them from another box? (which is properly configured)
1161093160 M * _are_ well, I have no idea where it takes the URLs from.
1161093228 M * _are_ unfortzunately errors scriolled away, will restart to get them again
1161093490 J * Piet hiddenserv@tor.noreply.org
1161093854 M * _are_ Retrieving GPG key from file:///etc/pki/rpm-gpg/RPM-GPG-KEY-fedora
1161093854 M * _are_ GPG key retrieval failed: [Errno 5] OSError: [Errno 2] No such file or directory: '/etc/pki/rpm-gpg/RPM-GPG-KEY-fedora'
1161093874 M * _are_ uhm, obvious as I have a debian system. now where can I get the fedora keys?
1161094071 M * _are_ found some in various directories on various servers in the net. Will just try one.
1161094321 Q * Radiance Ping timeout: 480 seconds
1161097186 J * Rich_Estill ~restill@c-24-11-195-139.hsd1.mi.comcast.net
1161097336 N * Bertl_zZ Bertl
1161097341 M * Bertl morning folks!
1161097365 M * Borg- evening
1161097368 M * Rich_Estill Hey B
1161097385 M * Bertl daniel_hozac: ah, cool, so it is a powerpc issue with dietlibc, I guess?
1161097396 M * daniel_hozac Bertl: it seems that way.
1161097407 M * Bertl for a moment I was worried :)
1161097425 M * Bertl daniel_hozac: did you get around playing with the more restrictive variant too?
1161097432 M * daniel_hozac _are_: ah! thanks! i'll look in to that.
1161097446 M * daniel_hozac Bertl: not yet, i'll do that now.
1161097520 M * Bertl okay, great! will see if I can locate the vlogin issue on powerpc, is there a way to disable the vlogin easily?
1161097547 M * daniel_hozac no :( you'll have to edit /usr/sbin/vserver.
1161098162 M * Bertl hmm, removing the --vlogin from the OPTS_VCONTEXT_ENTER should do it?
1161098228 M * daniel_hozac yep.
1161098242 M * doener what's the problem with vlogin?
1161098255 M * Bertl on powerpc, I get the following:
1161098263 M * Bertl - login seems to work fine, prompt
1161098276 M * Bertl - everything I type is lost, until I press CTRL-D
1161098287 M * _are_ daniel_hozac: I seem to have a fc5 vserver now
1161098289 M * Bertl - at this point, it is sent and executed
1161098317 M * Bertl I can confirm now, after disabling the --vlogin, everything seems to work fine (on enter)
1161098325 M * daniel_hozac _are_: great, i'll fix those file:/// URLs in svn.
1161098411 M * daniel_hozac Bertl: hmm, that sounds like select() not returning when there's available data.
1161098432 M * Bertl it seems to have to do with the 'new' pts or so
1161098481 M * Bertl btw, why does the fc6 package (source rpm) explicitely exclude ppc arch?
1161098510 M * daniel_hozac if you put WRITE_MSG(1, "."); at line 278 in src/vlogin.c, do you get a dot for each character you type?
1161098541 M * daniel_hozac or do you just get a stream of dots once you press ctrl+d?
1161098556 M * Bertl will check shortly, switching kernel atm
1161098558 M * daniel_hozac https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=172389
1161098567 M * daniel_hozac i think.
1161098658 M * Bertl hmm, yeah, very likely :)
1161098972 J * Radiance 1456fabc0f@halt.1984world.eu
1161099041 M * Bertl wb Radiance!
1161099049 M * Radiance thanks mate :)
1161099191 Q * mire Ping timeout: 480 seconds
1161099313 M * waldi Bertl: hmm, will bindmounting /dev read-only work, I think no
1161099319 M * Bertl yep
1161099426 M * waldi i had several developers now which killed their /dev and a read-only mount would be a good idea. they use chroots, which needs a bindmounted /dev and sometimes delete it without umounting them
1161099498 M * daniel_hozac hehe, how nice of them.
1161099510 M * mnemoc waldi: /dev as ro doesn't work
1161099535 M * Bertl mnemoc: why not? should work fine regarding device nodes?
1161099562 M * mnemoc at least /dev/null doesn't
1161099587 M * waldi you may need to explicitely allow the overwrite of the access modes of device nodes
1161099603 M * mnemoc i had to populate a tmpfs on pre-start :\
1161099609 M * Bertl mnemoc: sure that your /dev/null isn't accidentially a file 
1161099617 M * mnemoc sure
1161099629 M * daniel_hozac works fine here.
1161099660 M * daniel_hozac (mount -o ro --bind /dev /tmp/test; echo test > /tmp/test/null)
1161099677 M * waldi with a vserver kernel?
1161099690 M * Bertl same here on powerpc
1161099698 M * waldi hmm
1161099705 M * Bertl yes, definitely with a vserver kernel
1161099711 M * mnemoc i was trying to share an /dev between the guests because their root are nodev....
1161099719 M * doener I did just remount /dev ro, works as well
1161099722 M * doener (on vanilla
1161099727 M * doener 2.6.19-rc1-gcb1055fb
1161099728 M * mnemoc maybe it's different if it's a bind and not a "new partition"
1161099737 M * waldi will test
1161099771 M * Bertl mnemoc: both daniel_hozac  and I tried with bind mounts a few seconds ago
1161099774 M * waldi okay, works
1161099793 M * waldi it even works with a fresh tmpfs
1161099800 M * waldi root@wavehammer:/tmp/test# echo test > null 
1161099800 M * waldi root@wavehammer:/tmp/test# echo test > nul 
1161099800 M * waldi zsh: read-only file system: nul
1161099816 M * Bertl yep
1161099837 M * Bertl what I don't know is if pipes (fifos)work or not
1161099859 M * Bertl but devices are fine on ro partitions
1161099894 M * mnemoc uhm, now the test worked...
1161099908 M * mnemoc ah! no
1161099911 M * mnemoc /dev/log
1161099914 M * mnemoc not /dev/null
1161099915 M * mnemoc :D
1161099922 M * mnemoc sorry for the confusion
1161099929 M * daniel_hozac UNIX sockets likely won't work, no.
1161100152 J * stefani ~stefani@tsipoor.banerian.org
1161100210 M * Bertl hola stefani! early today?
1161100232 M * stefani Bertl:   have had coffee.
1161100285 M * Bertl ah, the RealThing(tm) not that inadequate substitute called 'sleep' :)
1161100316 M * mnemoc :D
1161100640 M * Bertl daniel_hozac: I don't get any dots until CTRL-D
1161100861 M * Bertl .[example:/]# ..exit
1161100861 M * Bertl logout
1161101622 J * bonbons ~bonbons@83.222.36.111
1161101651 M * doener Bertl: http://people.linux-vserver.org/~doener/vlogin-copy-settings.diff -- does that help?
1161101653 M * daniel_hozac Bertl: so select isn't returning.
1161101701 M * daniel_hozac doener: is it really necessary to reset that after every keystroke?
1161101744 M * daniel_hozac and shouldn't the cfmakeraw use termbuf?
1161101748 M * doener I have no clue (anymore), I just see that my a vlogin attempt I've written about half a year ago
1161101840 M * daniel_hozac difference between TCSAFLUSH and TCSADRAIN? the latter discards all current input/output?
1161101852 M * doener no, the former ;)
1161101886 M * doener my code actually uses TCSANOW, but TCSADRAIN seems more appropriate
1161101887 M * daniel_hozac hmm, that seems odd.
1161102106 M * daniel_hozac but what would that actually change? isn't the terminal_raw() call at the beginning of the vlogin process sufficient?
1161102125 M * doener that sets the terminals into an initial state
1161102144 M * doener the added calls copy terminal mode changes from the pty to the original terminal
1161102196 M * doener eg. xterm runs on pty1 vlogin provides pty2 to the vserver process
1161102216 M * doener if pty2 mode is changed, the new mode was never transferred to pty1
1161102264 M * daniel_hozac hmm. i believe there's a reason we don't do that.
1161102271 M * daniel_hozac IIRC that messed all sorts of things up.
1161102322 M * daniel_hozac and if it was a problem, things like password prompts wouldn't work, right?
1161102383 M * doener if I remove these calls from my program, then eg. vim produces garbage
1161102395 M * doener and otherwise it's pretty identical to the vlogin code
1161102445 M * daniel_hozac vim works fine for me.
1161102704 M * daniel_hozac how do you get the garbage? do you have to do something special?
1161102728 M * doener no, appears right after starting vim (with my tool, didn't check vlogin, still on .210)
1161102843 M * doener nvm...
1161102854 M * doener I didn't have an initial setup before the loop
1161102873 M * daniel_hozac ;)
1161102881 Q * yarihm Quit: Leaving
1161103666 Q * rob-84x^ Quit: That's it for today
1161103675 J * rob-84x^ rob@submarine.ath.cx
1161103691 M * Bertl wb rob-84x^!
1161103750 Q * Piet Ping timeout: 480 seconds
1161103967 J * bronson ~bronson@66.160.177.208
1161103975 Q * shedi Quit: Leaving
1161104586 M * Bertl okay, have to run now .. back later!
1161104592 N * Bertl Bertl_oO
1161104917 J * Osgiliath ~abigor@kurzweil.no-ip.org
1161105111 M * Osgiliath Hi, I'm trying to install Oracle 10g in a Debian vserver. The setup freezes each time it wants to start the Oracle Net Listener on port 1521.
1161105123 M * Osgiliath Does anyone know how to do this ?
1161105133 M * Osgiliath I'm running on 2.6.17.13-vs2.0.2.1
1161105487 J * Nei ~ailin@85.14.218.84
1161105513 M * Nei yo
1161105564 M * Nei what would I need to do to permit a vserver to mount a truecrypt image? the specific message is 'truecrypt: No free loopback device available for file-hosted volume' so I guess I need to make a loop device available somehow?
1161105898 J * hijacker ~hijacker@cable-84-43-140-185.mnet.bg
1161105903 M * hijacker hi all
1161105972 M * hijacker i have that problem, where when one specific user logs in to the system via ssh, he gets that error message:
1161105975 M * hijacker No mail.
1161105975 M * hijacker -bash: /dev/null: Permission denied
1161105975 M * hijacker -bash: /dev/null: Permission denied
1161105975 M * hijacker -bash: /dev/null: Permission denied
1161105983 M * hijacker and some more lines of that
1161105991 M * hijacker i checked /dev/null
1161105999 M * hijacker it seems to be crw-r--r--
1161106021 M * hijacker the direcotry mounted under the host is with options: defaults
1161106029 M * hijacker which sets the dev option
1161106036 Q * ensc|w Ping timeout: 480 seconds
1161106070 M * hijacker directory = /vservers
1161106088 M * hijacker anyone might know where the problem lies?
1161106228 M * hijacker and thet's in the auth.log file:
1161106230 M * hijacker Oct 17 20:33:02 vn sshd[31291]: Accepted password for vlado79 from 192.168.0.6 port 1607 ssh2
1161106230 M * hijacker Oct 17 20:33:02 vn sshd[31293]: (pam_unix) session opened for user vlado79 by (uid=0)
1161106230 M * hijacker Oct 17 20:33:02 vn pam_limits[31293]: setrlimit limit #13 to soft=-1, hard=-1 failed: Operation not permitted; uid=0 euid=0
1161106230 M * hijacker Oct 17 20:33:02 vn pam_limits[31293]: setrlimit limit #14 to soft=-1, hard=-1 failed: Operation not permitted; uid=0 euid=0
1161106231 M * hijacker Oct 17 20:33:02 vn sshd[31294]: error: open /dev/tty failed - could not set controlling tty: Permission denied
1161107420 J * ensc|w ~ensc@www.sigma-chemnitz.de
1161108104 M * daniel_hozac Nei: sounds like it.
1161108146 M * daniel_hozac hijacker: /dev/null should be 0666.
1161108180 M * mnemoc defaults sets nodev for me, not dev
1161108187 M * daniel_hozac mnemoc: as it should.
1161108202 M * mnemoc [19:29:09] <hijacker> the direcotry mounted under the host is with options: defaults
1161108205 M * mnemoc [19:29:18] <hijacker> which sets the dev option
1161108219 M * mnemoc daniel_hozac: fully ack, i was commenting ----^
1161108258 M * daniel_hozac depends on which fstab, i guess.
1161108356 M * daniel_hozac and which filesystem it's referring to.
1161108367 M * mnemoc :)
1161110208 J * rgl Rui@217.129.151.190
1161110211 M * rgl hello
1161110326 M * hijacker reiserfs
1161110344 M * hijacker the man mount tells that is dev
1161110356 J * coocoon ~coocoon@dslb-084-057-217-171.pools.arcor-ip.net
1161110369 M * coocoon hello
1161110374 M * hijacker  defaults
1161110374 M * hijacker                      Use  default  options: rw, suid, dev, exec, auto, nouser,
1161110374 M * hijacker                      and async.
1161110407 M * hijacker daniel_hozac, i cannot chmod 0666 /dev/null
1161110420 M * mnemoc hijacker: mount it with nodev
1161110458 M * mnemoc err, with dev
1161110663 M * hijacker mnemoc, according to the masnual it is already with dev mounted
1161110669 M * hijacker *manual
1161110691 M * mnemoc ok
1161110877 M * hijacker hmm
1161110883 M * hijacker problem resolved
1161110893 M * hijacker chown root.root /dev/null
1161110900 M * hijacker then chmod 0666 /dev/null
1161110915 M * hijacker /dev/null became 0666 now ;_)
1161110951 Q * coocoon Quit: KVIrc 3.2.0 'Realia'
1161110955 M * hijacker also changed /dev/tty to 0666
1161111316 J * shedi ~siggi@inferno.lhi.is
1161112179 M * rgl is the libvserver and vcc read to use?
1161112189 M * rgl err s,read,ready,
1161112246 M * rgl or should one use vserver-utils?
1161112326 M * daniel_hozac i assume you mean util-vserver?
1161112368 M * rgl oh yes.  I confuse the two :(
1161112409 M * daniel_hozac i think vcd/vstatd/vwrappers/libvserver should work, but i haven't gotten around to testing it.
1161112450 M * rgl is Hollow arround?
1161112479 M * daniel_hozac not right now.
1161112503 M * rgl in the last day or so, I never seen him here
1161112526 M * daniel_hozac i think he should have internet access at home again from tomorrow.
1161112527 M * rgl but I guess I'll just use util-vserver 
1161112552 M * rgl ah cool :D
1161112584 M * daniel_hozac anything in particular wrong with util-vserver? :)
1161112660 M * rgl see http://svn.linux-vserver.org/viewvc.py/libvserver/trunk/README
1161112778 M * daniel_hozac and?
1161112797 M * daniel_hozac IMHO the API compatibility is a good thing, not a bad one :)
1161112829 M * rgl I'm more woried about "though making it quite hard to fix bugs or add new
1161112830 M * rgl features  in an adequate timeframe" :D
1161112852 M * daniel_hozac which bugs does it have? :)
1161112885 M * rgl I dunno, but for him to say that, I guess he was bitten for some
1161112978 M * daniel_hozac not that i can remember.
1161113291 M * rgl why is dietlibc a "recommended build environment"
1161113300 M * rgl ?
1161113328 M * daniel_hozac because glibc can load libraries from the guest.
1161113456 M * rgl I didn't get the relation :|
1161113964 J * hmeyer ~hmeyer@p54BDA0CD.dip0.t-ipconnect.de
1161113989 M * daniel_hozac that means glibc can be a security risk. thus dietlibc, which doesn't do that, is recommended.
1161114055 M * hmeyer Hi...anybody out there using Ubuntu as VServer host?
1161114110 M * hmeyer At least in Dapper I am not able to apply the kernel-patch-vserver.... (Micah...whats wrong there?)
1161114147 M * daniel_hozac kernel-patch-vserver is ancient stuff.
1161114155 M * rgl daniel_hozac, ah ok. 
1161114166 M * rgl daniel_hozac, btw, known a rpm for util-vserver?
1161114190 M * daniel_hozac for what distribution? rpmbuild -tb util-vserver-0.30.211.tar.bz2 should work as well.
1161114193 M * micah hmeyer: kernel-patch-vserver was made for debian kernels, I doubt it would apply to an ubuntu kernel, as they do different things to it. Also, that package has been removed from unstable (and soon will be removed from etch), so it is quite outdated now
1161114229 M * hmeyer oh..ok, you should remove it from dapper as well then
1161114250 M * hmeyer micah: but whats the way to vserver for ubuntu users then?
1161114322 M * daniel_hozac IIRC there was someone patching Ubuntu kernels.
1161114328 M * daniel_hozac they were linked from some Ubuntu wiki.
1161114391 M * Borg- micah: hmmm?? for Debian Kernel?
1161114406 M * Borg- AFAIK debian uses unpatched Linux Kernel...
1161114428 M * Borg- Suse.. RedHat.. they patch Linux Kernel.. and its unusable for such purposes
1161114441 M * daniel_hozac "such purposes"?
1161114451 M * Borg- daniel_hozac: for wider OpenSource I would say.
1161114460 M * daniel_hozac huh?
1161114490 M * Borg- sometimes you just cannot apply the kernel patches.. because they collide w/ redhat/suse ones
1161114502 M * daniel_hozac well, yeah.
1161114571 M * doener same is often true for any two patches you might want to apply
1161114579 M * micah hmeyer: I dont maintain the package in ubuntu
1161114598 M * doener if you don't want the patches your vendor has applied, "compose" one yourself
1161114611 M * micah Borg-: no, debian patches the vanilla kernel
1161114620 M * doener (well, that didn't work for the RedHat 2.4+NPTL kernels)
1161114628 M * micah Borg-: as does redhat, and ubuntu
1161114647 M * hmeyer micah (or all): is there the one right way to patch the ubuntu kernel?
1161114650 M * Borg- micah: huh? didnt know that..
1161114665 M * Borg- at least now I understand why I dont like debian ;)
1161114675 M * daniel_hozac micah: is that still true? waldi said Debian didn't do patches anymore.
1161114680 M * micah hmeyer: I'm sorry I dont know anything about ubuntu's kernels
1161114696 M * micah daniel_hozac: well the debian kernel has the vserver patch... :)
1161114705 M * hmeyer who might know?
1161114709 M * daniel_hozac micah: right, but that was the only one.
1161114712 M * daniel_hozac hmeyer: https://help.ubuntu.com/community/VServer
1161114713 M * Borg- daniel_hozac: hope vserver uses linux.org kernel to do patch.. not any distro.. right?
1161114714 M * doener Borg-: hm, do you like any binary distro?
1161114727 M * Borg- doener: yeah.. Slackware :)
1161114732 M * micah Borg-: yes, linux-vserver uses vanilla kernel
1161114740 M * rgl daniel_hozac, humm we need g++ for util-vserver?
1161114760 M * hmeyer daniel: there's just a link to a ready made kernel-archive, but no outline how the author got there
1161114772 M * daniel_hozac rgl: possibly, but i think the last time we looked in to that, it turned out it wasn't needed.
1161114774 M * waldi daniel_hozac: we patch the kernel, but we try to limit the number of patches
1161114779 M * micah daniel_hozac: well, I know that its pretty much impossible to maintain the security of a kernel once it has frozen without applying patches cherry-picked from upstream fixes
1161114794 M * rgl daniel_hozac, so you can remove that check from configure :)
1161114798 M * rgl daniel_hozac, pulease :)
1161114826 M * daniel_hozac i suppose. but g++ is really basic development tools...
1161114849 M * rgl daniel_hozac, its more cruft that one needs to have :(
1161114867 M * daniel_hozac rgl: you don't build on your servers, do you?
1161114889 M * rgl daniel_hozac, I will, because I have no 64 bit box :(
1161114952 M * rgl daniel_hozac, but whats the problem building there?
1161114979 M * daniel_hozac rgl: you end up with lots of unneeded things installed :)
1161114986 M * daniel_hozac (i guess that's where guests are handy :))
1161114986 Q * hmeyer Quit: Verlassend
1161115006 M * rgl daniel_hozac, ah yeah, thats for sure.  but I intend to have a build guest like you say :D
1161115017 M * rgl but still, I have to bootstrap it :D
1161115105 M * doener daniel_hozac: what's the problem with removing the few required packages after building the tools?
1161115117 M * daniel_hozac doener: none :)
1161115132 M * doener it's just 4 or 5 packages IIRC
1161115169 M * doener the rest are dependencies that aptitude will remove automatically (if you use aptitude that is ;)
1161115557 Q * lilalinux_ Remote host closed the connection
1161115577 M * Borg- aaaaaaaaargh 
1161115589 M * Borg- wasted 2:30 hour.. because of stupid app
1161115616 M * Borg- I hate dumb developers.. he wrote in his l33t scripts: /usr/local/bin/python
1161115620 M * Borg- fucking smart.. 
1161115676 M * Borg- anyway.. good nite
1161115718 M * meandtheshell http://google.com/trends?q=vserver%2C+linux+vserver%2C+xen%2C+openvz%2C+qemu&ctab=1&geo=all&date=all well, not a world moving thing but also not uninteresting ... 
1161116401 M * doener "linux vserver" probably covers a lot of hosting offers
1161116491 M * doener ehrm, s/linux "//
1161116500 M * doener I give up...
1161116503 J * debugger Rui@217.129.151.190
1161116602 M * meandtheshell doener: I guess we understood :)
1161116626 M * meandtheshell doener: I think that assumption is correct
1161116966 Q * bonbons Quit: Leaving
1161116967 Q * rgl Ping timeout: 480 seconds
1161117215 Q * hijacker Quit: Leaving
1161117722 Q * FireEgl Ping timeout: 480 seconds
1161118532 J * Aiken ~james@tooax7-199.dialup.optusnet.com.au
1161118686 Q * _are_ Ping timeout: 480 seconds
1161118982 Q * debugger Ping timeout: 480 seconds
1161119529 J * mire ~mire@40-166-222-85.adsl.verat.net
1161119567 J * debugger Rui@217.129.151.190
1161119594 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1161119605 Q * cdrx 
1161119648 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1161120055 J * _are_ ~are@62.112.159.81
1161120253 Q * cdrx Quit: Leaving
1161120308 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net
1161120375 P * stefani I'm Parting (the water)
1161120397 Q * cdrx 
1161120716 N * debugger rgl
1161120762 M * rgl do we need the e2fsprogs package to build util-vserver rpm?
1161120793 M * daniel_hozac yes.
1161121099 M * rgl daniel_hozac, but why?  it build fine here without it :/
1161121179 M * daniel_hozac i don't know. i'm sure there's a reason for it :)
1161121848 Q * mnemoc Quit: leaving
1161121867 J * mnemoc ~amery@kilo105.server4you.de
1161122087 Q * michal` Ping timeout: 480 seconds
1161122346 J * michal` ~michal@www.rsbac.org
1161122798 Q * dna_ Quit: Verlassend
1161122897 Q * meandtheshell Quit: exit (0);
1161122916 Q * mire Ping timeout: 480 seconds
1161123375 M * rgl error: Failed dependencies:
1161123375 M * rgl         init(util-vserver) is needed by util-vserver-0.30.211-0.i386
1161123380 M * rgl daniel_hozac, any ideia about that?
1161123388 M * rgl what init is that?
1161123411 M * rgl the .spec file has this: Requires:       init(%name)
1161123691 Q * Rusty` 
1161123702 Q * fosco Ping timeout: 480 seconds
1161123706 J * debugger Rui@217.129.151.190
1161123744 N * Bertl_oO Bertl
1161123749 M * Bertl evening folks!
1161124013 N * debugger rgl_
1161124020 M * rgl_ hey :)
1161124172 Q * rgl Ping timeout: 480 seconds
1161124177 N * rgl_ rgl
1161124202 M * rgl Bertl, o you have any ideia about that rpm error when I try to install util-vserver?
1161124310 M * Bertl IIRC, you have to install the core/libs/sysv and such stuff
1161124397 M * rgl I also though so, but sysv needs util-vserver :(
1161124408 M * rgl so it borks the install... maybe I should force it :|
1161124420 M * Bertl nope, try to specify all of them on a single line
1161124431 M * Bertl rpm -U ....
1161124518 M * rgl oh, I've forced it already and it worked
1161124569 M * rgl the package doesn't seem right, because there is a recursive dependency between sysv and normal util-vserver
1161124637 M * Bertl hmm, I think those are allowed and usually get resolved quite fine by installing both packages at once
1161124826 M * rgl odd, the rpm package does not unstrip the binaries :|
1161125062 M * Bertl unstrip?
1161125126 J * derjohn2 ~aj@dslb-084-058-200-120.pools.arcor-ip.net
1161125184 M * rgl Bertl, I mean, they are not strip'ed
1161125279 M * Bertl ah, well, should be, at least they are diet binaries
1161125293 M * Bertl i.e. they are a few k byte typically
1161125670 M * rgl oh, vserver-stat does not run, it says something about being unable to open /proc/uptime, but I can cat it OK :(
1161125695 M * derjohn2 Bertl, did you get updates from ntrs about the mysql prob? I have something on my machine, that might interesting. Maybe not evenvserver realted, but maybe it is:http://paste.linux-vserver.org/501 
1161125709 M * Bertl rgl: you probably didn't run the vprocunhide script yet
1161125731 M * Bertl rgl: vserver-stat uses the spectator context (xid=1) which is probably still disabled
1161125755 M * rgl Bertl, ah, sounds like it then.  but where is that script?
1161125756 M * Bertl derjohn2: ntrs said: issues are gone now (basically)
1161125777 M * rgl Bertl, /etc/init.d/vprocunhide
1161125778 M * rgl  ?
1161125791 M * Bertl sounds good
1161125798 M * derjohn2 ah, fine.... I just builfing my rc40 kernel
1161125824 M * derjohn2 Bertl, did you already get feedback about the new init ?
1161125824 M * rgl Bertl, that was it. thx :))
1161125843 M * Bertl derjohn2: well, we will drop it for now ... too many issues
1161125852 M * Bertl rgl: you're welcome!
1161125973 M * derjohn2 Bertl, meaning I should wait for rc41 ?
1161126025 M * Bertl you could test drive 2.0.2.2-rc1 in the meantime
1161126044 M * Bertl should have rc41 ready in an hour or so
1161126055 J * debugger Rui@217.129.151.190
1161126069 M * derjohn2 Bertl, you mean /me ? I cannot use stable as long as it doesnt provide capability masking.
1161126080 M * derjohn2 rc41 is cool .:)
1161126083 M * Bertl i.c. :)
1161126097 M * derjohn2 Bertl, did we every each an rc level > 99 ?
1161126151 P * Rich_Estill Leaving
1161126174 M * derjohn2 *ever
1161126199 M * derjohn2 so the rc41 will be like rc40 but without the new init ?
1161126214 N * debugger rgl_
1161126247 M * Bertl derjohn2: yes, basically, plus a few minor fixes
1161126315 M * derjohn2 kk, i'll wait then!
1161126433 Q * Osgiliath Quit: The Sign Of Wrath Awaked
1161126487 Q * rgl Ping timeout: 480 seconds
1161127305 M * Bertl daniel_hozac: any results regarding the pid_task() alternatives?
1161127706 Q * rob-84x^ Ping timeout: 480 seconds
1161127976 M * rgl_ 'VERIFYCAP' can be executed as root only
1161127976 M * rgl_ capabilities are not enabled in kernel-setup
1161127981 M * rgl_ what is that error? :|
1161127994 M * rgl_ it happens when I run, vserver example start
1161128013 M * Bertl that sounds strange, as linux-vserver usually ensures that the capability system is compiled into the kernel
1161128014 M * rgl_ I forgot to enable something the the host kernel?
1161128032 Q * dmax Ping timeout: 480 seconds
1161128055 M * Bertl rgl_: the first message states that you are not root?
1161128073 M * Bertl could it be, that you aren't or that you are inside a guest?
1161128074 M * rgl_ Bertl, you known which is the .config entry that enables capabilities?
1161128084 M * rgl_ Bertl, nope. I'm root, and on the host
1161128103 M * Bertl CONFIG_SECURITY_CAPABILITIES=y
1161128130 M * Bertl maybe some se-linux or similar activated?
1161128153 M * rgl_ nope.  I've build the vanilla kernel by hand, and no selinux
1161128167 M * rgl_ I have no SECURITY_CAPABI in the .config file :|
1161128177 Q * s0undt3ch Ping timeout: 480 seconds
1161128184 M * Bertl how did you manage to do so?
1161128211 M * rgl_ no ideia *G*
1161128224 M * rgl_ in what menu is that stuff?
1161128247 M * Bertl the only idea I have is that you copied the .config there an didn't do a 'make menuconfig' or 'make oldconfig'
1161128260 M * rgl_ I did both *G*
1161128275 M * Bertl what kernel/patch version do you use right now?
1161128275 M * rgl_ but indeed, the menu doesn't have the CAPABILITI active!
1161128289 M * Bertl it should not be selectable at all
1161128298 M * rgl_ [*] Enable different security models
1161128303 M * rgl_ ---   Default Linux Capabilities
1161128304 M * Bertl i.e. it should be auto selected by linux-vserver
1161128313 M * rgl_ the first one was disabled
1161128325 M * rgl_ (the Enable different security models)
1161128333 M * Bertl should be fine too, because without that it's the default
1161128345 M * rgl_ huh?
1161128366 M * Bertl i.e. capabilities are the default if you disable different security models
1161128368 M * rgl_ the "Default Linux Capabilities" only happear after I select " Enable different security models"
1161128377 M * rgl_ ah!
1161128406 M * rgl_ anyways, I have the CONFIG_SECURITY_CAPABILITIES=y on the config :)
1161128415 M * rgl_ anyways, I *now* have...
1161128521 M * Bertl could you run the testme.sh on the 'current' kernel
1161128536 M * Bertl and upload the output to paste.linux-vserver.org for me?
1161128709 M * rgl_ Bertl, the rpm build has run the test suite, but maybe because it run as non-root all tests haved pass
1161128728 M * rgl_ http://paste.linux-vserver.org/502
1161128759 M * rgl_ Bertl,  oh... w8, what testme.sh?
1161128782 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/
1161128789 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh
1161128928 M * rgl_ Bertl, oh well, it doesn't run with the new kernel either :(
1161128950 M * Bertl that's what I thought, what kernel and what patches?
1161129007 M * rgl_ 2.6.17.13 and patch-2.6.17.13-vs2.0.2.1
1161129026 M * rgl_ Bertl, should I run testme.sh as root, right?
1161129035 M * Bertl yes, as root on the host
1161129101 M * rgl_ Bertl, http://paste.linux-vserver.org/503
1161129105 M * rgl_ all seem fine
1161129124 M * Bertl okay, looks good ...
1161129140 M * Bertl except for two things:
1161129175 M * Bertl it seems your tools have been compiled with legacy interfaces disabled
1161129216 M * Bertl (could cause problems, but might be fine too, you have to talk to daniel_hozac or ensc for details, I've lost the overview :)
1161129246 M * Bertl no idea what the second thing is :)
1161129319 J * s0undt3ch ~s0undt3ch@81.193.56.104
1161129329 Q * s0undt3ch 
1161129345 J * s0undt3ch ~s0undt3ch@81.193.56.104
1161129362 M * Bertl wb^2 s0undt3ch! :)
1161129377 M * rgl_ I've rechecked the kernel config, and it seems ok its at  http://paste.linux-vserver.org/504
1161129383 M * s0undt3ch thanks
1161129423 M * Bertl rgl_: yep, kernel side is fine (as it seems)
1161129452 M * Bertl rgl_: the fact that testme.sh succeeds, shows that the capabiltiy system is active
1161129477 M * ntrs Bertl, I have a guest complaining of the mysql issues. The host currently runs 2.6.17.13-211rc35
1161129495 M * ntrs I will try a newer version and we'll see what happens. This is on an x86 system.
1161129512 M * Bertl okay, good, how soon do you want to try?
1161129530 M * Bertl (jsut asking, because rc41 should be out shortly)
1161129555 M * ntrs what is in rc40 and 41?
1161129568 M * ntrs the init fix?
1161129573 M * Bertl the fake init is fixed and the signalling should be final