1160179209 N * hardwire-afk hardwire
1160179247 J * Hurga nobody@p508AA90E.dip0.t-ipconnect.de
1160180217 J * mire_ ~mire@76-167-222-85.adsl.verat.net
1160180268 Q * DreamerC_ Quit: leaving
1160180284 J * DreamerC ~dreamerc@61-217-246-141.dynamic.hinet.net
1160180687 J * Thorsten ~Thorsten@dslb-084-058-144-091.pools.arcor-ip.net
1160180691 N * hardwire hardwire-afk
1160180703 M * Bertl daniel_hozac, doener: could you have a look at the switch.c and check the following conclusions:
1160180780 M * Bertl - we have two different compat cases for 32bit mapping
1160180813 M * Bertl - one is the __COMPAT when CONFIG_COMPAT, as used in VCMD_add_dlimit
1160180845 M * Bertl - the other is a special __COMPAT_IA32 when CONFIG_IA32_EMULATION and CONFIG_COMPAT
1160180867 M * Bertl  (as used by VCMD_get_rlimit)
1160180925 M * Thorsten Would it be a very silly idea to copy a vserver with cp -ax? 
1160180940 M * Bertl daniel_hozac, doener: the second one is caused by the unfortunate alignment of the structures
1160180962 M * Bertl Thorsten: not at all, although I'd prefer rsync/dump for several reasons
1160180993 M * Thorsten The problem is, I get:    vcontext: chroot(): Permission denied
1160181016 M * Thorsten According to google, this has been asked once
1160181037 M * Bertl and what was the conclusion?
1160181041 M * Thorsten found it at last. The directories of the vservers had the barrier attribut
1160181041 M * Thorsten set on all subdirectories
1160181066 M * Bertl well, but that should not be caused by a cp
1160181084 M * Bertl unless it is a broken cp or a very old kernel version
1160181099 M * Thorsten 2.6.18-vs2.0.2.1-t8-gbf5348d4-dirty :-P
1160181112 M * Bertl doesn't look very old to me :)
1160181114 M * doener the inheritance of +t happened on 1.25 only ;)
1160181116 M * Thorsten ;-)
1160181141 M * Bertl so where did the barrier come from in your case?
1160181149 M * Thorsten Can I track down which directory is the problem?
1160181165 J * marl ~matt@84.92.193.226
1160181173 M * Bertl ah, that was the result, not your findings, right?
1160181182 M * Bertl welcome marl!
1160181189 M * Thorsten me? Yes, that was the google result
1160181193 Q * Borg- Ping timeout: 480 seconds
1160181194 M * marl hi Bertl, how r u?
1160181209 M * Bertl marl: well, ill (flu), but otherwise fine
1160181218 M * Bertl Thorsten: okay, xid tagging?
1160181224 M * Thorsten I tried  bash -x vserver ubuntuX start  and the last command is:
1160181225 M * Thorsten exec /usr/sbin/vnamespace --new -- /usr/sbin/vserver ----nonamespace ubuntuX start
1160181232 M * Thorsten Bertl, not that I know
1160181237 M * marl i remember reading someware of a script that will copy a live server into a vserver, but i cant find it now :( anyone point me in the rite direction?
1160181247 M * marl how long u had the flu?
1160181253 M * Bertl Thorsten: okay, check for barrier flags on the guest dirs, just to make sure
1160181262 M * Bertl marl: a few days now ...
1160181272 M * marl proberly nearly over it then :)
1160181285 M * Thorsten how do I check this? lsattr?
1160181307 M * Bertl showattr would be best
1160181370 M * Thorsten Ah, they are different. The original was ---bui- ubuntu the copy has ---Bui- ubuntuX
1160181455 M * doener "interesting"
1160181558 M * Thorsten He he, works :-)
1160181716 M * marl ah found it now, 'virtulize' :)
1160181822 M * marl sounds like a daft question, but is it posible to start a vserver with networking, but not allow it access to the internet? im wanting to backup a server to a vserver, alter the config iles to allow it to run normally, and try and start it, without causing problems for the old server until im ready to change the boxes over
1160181898 M * Bertl yes, iptables is your friend
1160182007 M * marl :( thought u mite say that LOL
1160182034 A * Hurga would have said that, too.
1160182079 M * marl is it safe to rsync dev on a normal server into a vserver, or should i exclude it from the backup?
1160182255 M * Bertl normal dev into a guest will reduce security there
1160182268 M * Bertl i.e. guest dev is a reduced version 
1160182293 M * Bertl so basically I'd exclude dev and tmp
1160182310 M * marl ok thanks :)
1160182329 M * Bertl np, also --numeric-ids and -H is a good idea for rsync
1160182395 M * Hurga Bertl: not sure where I got my  /dev... where can I get the vserver reduced version?
1160182445 M * Bertl just create a new skeleton and take it from there
1160182551 A * Hurga looks and has only 17 entries in /dev. That probably will be the reduced version. :)
1160182672 M * Hurga Bertl: I remember to have seen a paper on chroot and vserver security somewhere... or maybe it was a presentation. Can't find it anymore, do you happen to know where it might be?
1160182857 M * Bertl phew, google is your friend there, I guess
1160182875 M * Bertl we had a few links on the old wiki ..
1160182885 M * doener Hurga: reduced dev is like 8 entries
1160182969 M * doener Hurga: try googling for enrico scholz, I guess you mean his presentation
1160183174 M * doener Hurga: http://www.13thfloor.at/~ensc/util-vserver/files/doc/virtual-servers.pdf#search=%22enrico%20scholz%20vserver%20chroot%22
1160183177 M * doener that one?
1160183351 M * Hurga Yes! excellent. Thank you! I was only finding http://infotage.pf-lug.de/2006/downloads/vortraege/vserver.pdf#search=%22%22enrico%20scholz%22%20vserver%20security%22
1160183443 M * doener Bertl: btw, was the alignment discovery an answer to your compat question that you gave yourself, or shall I still tell you that I confirmed your conclusions? (which I hereby do ;)
1160183503 Q * ntrs Quit: Leaving
1160183524 M * Bertl doener: yeah, seems I required a refresh on that
1160183550 M * Bertl maybe we should add a comment there or do a special __COMPAT_IA32 for that
1160184302 Q * Thorsten Quit: Leaving
1160186296 Q * samuel Ping timeout: 480 seconds
1160187266 Q * weeble Ping timeout: 480 seconds
1160187275 Q * ensc Killed (NickServ (GHOST command used by ensc_))
1160187284 J * ensc ~irc-ensc@p54B4FD59.dip.t-dialin.net
1160187434 Q * Piet Write error: connection closed
1160187508 J * weeble ~weeble@81.52.144.1
1160187641 J * Piet hiddenserv@tor.noreply.org
1160187976 M * Bertl daniel_hozac, doener: still awake?
1160188110 M * Bertl here is the first version of the kernel interface (untested)
1160188122 M * Bertl http://vserver.13thfloor.at/Experimental/delta-smon-feat01.diff
1160188893 M * doener been doing some tests how to get data to the disk fast... until I got distracted during some google search
1160188956 M * Bertl well, the userspace (and then disk) transfer is straight forward IMHO, your part is to interpret the data somehow :)
1160189019 M * Bertl anyway, I guess it's enough for me for today .. would appreciate a check of the code though, will write the reader tomorrow and test it on a running kernel
1160189027 M * marl whats the format of entrys in bcapabilities ? im wanting to enable cap_sys to allow bind9 to run within my guest? (i know its a bad idea, but its a temp fix, till ht emachine can be moved to a non-caps bind)
1160189061 M * doener hm id % NR_CPUS.... shouldn't id >= NR_CPUS return -EINVAL?
1160189079 M * Bertl doener: yeah, probably a better choice
1160189104 M * Bertl marl: just add the cap entry there (cap_sys is no valid entry)
1160189115 J * s0undt3ch_ ~s0undt3ch@bl7-242-167.dsl.telepac.pt
1160189119 M * marl sorry was shortning it :)
1160189143 M * Bertl e.g. echo CAP_SYS_ADMIN >bcapabilies
1160189144 M * marl so all i would need is : CAP_SYS_RESOURCE in the file?
1160189155 M * marl ah thanks :)
1160189235 M * marl :( still getting capset failed, operation not permitted from bind
1160189240 M * doener guess I'll head to bed too... will review the patch later
1160189250 M * doener marl: did you use CAP_SYS_RESOURCE?
1160189265 M * marl nope, CAP_SYS_ADMIN
1160189272 M * doener thought so ;)
1160189279 M * doener that was just an example how to do it
1160189305 M * Bertl okay, have a good one then ... I'm off to bed ...
1160189307 M * marl lol, so i was rite the first time :)
1160189317 M * marl night Bertl, thanks for your help :)
1160189318 M * doener night Bertl!
1160189323 N * Bertl Bertl_zZ
1160189371 M * marl :) working :) thanks again
1160189551 Q * s0undt3ch Ping timeout: 480 seconds
1160189551 N * s0undt3ch_ s0undt3ch
1160191343 Q * Hurga Remote host closed the connection
1160191662 J * samuel ~samuel@72.18.248.90
1160191665 N * samuel samuel__
1160194296 Q * _are_ Ping timeout: 480 seconds
1160194946 Q * weeble Ping timeout: 480 seconds
1160195022 Q * samuel__ Ping timeout: 480 seconds
1160195179 J * weeble ~weeble@81.52.144.1
1160197086 J * samuel__ ~samuel@72.18.248.90
1160200132 Q * samuel__ Ping timeout: 480 seconds
1160200976 M * Loki|muh why does a kill 1 fail and a pkill init not?
1160200997 M * Loki|muh because of the number 1 which is the hosts init?
1160201016 M * Loki|muh which init is killed, when pkill init is called?
1160201045 M * Loki|muh I ask because there are problems in the upstart.postinstall, which is the new init-system of ubuntu
1160201054 M * Loki|muh there is done a kill 1
1160201300 M * essobi_ Heh
1160201771 J * meandtheshell ~markus@85-124-233-141.work.xdsl-line.inode.at
1160202102 J * tydol ~tydol@202.131.180.115.cmts5.guam.net
1160202115 M * tydol anyone around?
1160205588 M * ruskie hmm is it possible to run X in a vserver and how?
1160206711 M * ray6 X Applications or an X Server? X Applications should be no problem
1160206837 Q * Piet Quit: :tiuQ
1160207096 M * ruskie ray6, both
1160207694 M * Loki|muh ah, pkill itself doesnt find a init, but it does not return a error
1160207911 M * tydol basically, what does a vserver do?
1160207924 M * tydol i'm sorta new to it.
1160207978 M * ray6 it's like a better chroot :)
1160208013 M * ray6 you can jail a whole distribution into it and give it limited access on ressources, like it's own IP and so on, even to the root user inside
1160208292 M * tydol i don't see how this is useful.
1160208308 M * tydol could you give a useful example of what it can do?
1160208334 M * ray6 you could run a debian and a redhat server on the same machine on 2 different IPs
1160208365 M * ray6 or run multiple debian servers for multiple persons, while everyone is "root" on his machine he can't get into the other's
1160208373 M * tydol so it's like an operating system that manages other operating systems.
1160208412 M * ray6 there's only one linux kernel running in the host - that's different from Virtualization as done by Xen, VMWare, UML and so on
1160208420 M * ray6 it's less ressource intensive
1160209047 J * Borg- borg@cube.benet.uu3.net
1160209520 J * samuel__ ~samuel@72.18.248.90
1160209707 J * dna_ ~naucki@p54BCE138.dip.t-dialin.net
1160209802 J * dna___ ~naucki@p54BCF83B.dip.t-dialin.net
1160210211 Q * dna_ Ping timeout: 480 seconds
1160210849 J * bonbons ~bonbons@83.222.36.111
1160211084 J * dna_ ~naucki@p54BCF83B.dip.t-dialin.net
1160211311 M * h01ger moin - the last line in /proc/virtual/<xid>/sched is not explained on http://oldwiki.linux-vserver.org/HowTo+Read+ProcFS - from past questions i guess its usercpu, syscpu and ??? 
1160211321 Q * dna___ Ping timeout: 480 seconds
1160211865 J * Viper0482 ~Viper0482@p54977ED7.dip.t-dialin.net
1160212713 M * daniel_hozac h01ger: and on hold.
1160212786 M * daniel_hozac Loki|muh: does your guest run an init?
1160212799 M * daniel_hozac Loki|muh: i.e. are you using the plain initstyle for it?
1160212877 M * Loki|muh daniel_hozac: uhm, I think I'm using the default init-style. Don't know which one this is
1160212884 M * daniel_hozac sysv.
1160212894 M * daniel_hozac which means there is no init.
1160212920 M * Loki|muh ah, ic
1160212936 M * Loki|muh so its the hosts init which is shown in ps?
1160213024 M * daniel_hozac yep, it blends through.
1160213031 M * daniel_hozac but it's untouchable by the guest.
1160213194 M * Loki|muh ah ok
1160213545 M * h01ger daniel_hozac, on hold?
1160213559 M * daniel_hozac h01ger: yes, where the token bucket kept the guest from getting scheduled.
1160213642 M * h01ger ic
1160213665 Q * samuel__ Ping timeout: 480 seconds
1160213681 Q * ierk Quit: Leaving
1160214847 M * mugwump_ hmm, guess that etch fai feature needs some midnight oil, huh...
1160215059 Q * m4z Quit: brb
1160215139 J * m4z m4z@bastard-operator.from-hell.net
1160215142 N * mugwump_ mugwump
1160215185 M * mugwump well, I'll re-test head, but I can't promise to make the features into the etch FAI release
1160215646 Q * cehteh Remote host closed the connection
1160215738 J * cehteh ~ct@pipapo.org
1160215904 M * daniel_hozac mugwump: well, does the build method work on sarge?
1160215948 M * mugwump just had a successful build, yes
1160215995 M * daniel_hozac ok, thanks, that's what i wanted to know :)
1160216018 M * mugwump although, the shell at the end without -a is probably confusing
1160216028 M * mugwump it should at least have a message; it was mostly a debugging thing
1160216086 M * daniel_hozac so maybe we should make -a the default?
1160216089 M * mugwump hmm, also, if I press Ctrl+C at the prompt, it doesn't clean up the interface
1160216116 M * daniel_hozac hmm, it doesn't?
1160216119 M * mugwump Yes, probably make the -a switch work the other way; given none of the other build methods prompt
1160216136 M * mugwump no, the second time, I get the RTNETLINK answers: File exists
1160216140 M * mugwump (before the prompt appears)
1160216149 M * daniel_hozac oh, if you press ctrl+c and the "Press ENTER to begin..." prompt?
1160216153 M * daniel_hozac s/and/at/
1160216159 M * mugwump yeah
1160216172 M * daniel_hozac yeah, the trap is setup to late...
1160216181 M * daniel_hozac or rather, the interfaces enabled too early.
1160216201 M * mugwump sure; the trap should probably be set up for all build methods, I guess
1160216219 M * mugwump when the interface is set up, the trap should remain until the process is complete
1160216259 M * daniel_hozac a trap is setup for all build methods, it's what cleans up the directories if it fails.
1160216295 M * daniel_hozac but most build methods don't enable the interfaces, nor start a semi-complete guest for installation.
1160216315 M * mugwump aha
1160216358 M * mugwump If I give you a fai .deb you can test this with, would that help?  You'd need a sarge host or chroot
1160216400 M * daniel_hozac well, were the any other problems? or does it seem to work fine now?
1160216486 M * mugwump interrupting the build is not cleaning out processes, but I suppose that's not so awful
1160216527 M * daniel_hozac hmm, it really should.
1160216585 M * daniel_hozac i even added dynamic context support for that.
1160216593 M * daniel_hozac (even though they shouldn't be used :))
1160216975 M * daniel_hozac could you try adding an echo to the if [ $? -eq 0 -a -d "/proc/virtual/$S_CONTEXT" ] clause, to make sure it's getting executed?
1160217137 M * mugwump and my /usr is still being remounted RO by this script for some reason
1160217306 M * mugwump ok, so... letting it finish normally, the interface is not torn down, but the hook is definitely called
1160217415 M * mugwump I'm also getting :
1160217416 M * mugwump save_ctxinfo: open("/var/run/vservers/ns"): File exists
1160217427 M * mugwump have to delete that file manually
1160217458 Q * tydol 
1160217485 M * daniel_hozac humm.
1160217497 M * mugwump you know, I wonder if it's killing itself
1160217512 M * mugwump because my debug message after the vkill -s 9 doesn't echo
1160217513 M * daniel_hozac could you try http://people.linux-vserver.org/~dhozac/p/uv/experimental/vserver-build.fai ?
1160217520 M * mugwump sure
1160217552 M * daniel_hozac do you have another debug message inside the if, like right after the then?
1160217602 M * mugwump sure, I put one before each vkill
1160217613 M * mugwump and they both printed, but not one just after the fi
1160217642 M * daniel_hozac hmm, maybe the vkill returned non-zero?
1160217707 M * mugwump correct
1160217717 M * mugwump it gave an error
1160217725 M * daniel_hozac so we need || : after those, right?
1160217732 M * daniel_hozac or disable the set -e
1160217738 M * mugwump probably ... yeah set +e
1160217852 M * mugwump ok, so running the build again with --force still dies due to the /var/run/vservers/xxx existing
1160217931 M * mugwump hmm, if I interrupt the build, I get a shell immediately, and the trap is still running in the background
1160217944 M * mugwump but when it wakes up, it doesn't get an S_CONTEXT
1160218015 M * daniel_hozac do you get a shell inside the context?
1160218037 M * mugwump well I didn't pass -a
1160218090 M * daniel_hozac ah.
1160218102 M * daniel_hozac oh of course
1160218105 M * mugwump it's not prompting me any more, nor leaving me with a shell.  so that worked; I'll try with -a
1160218108 M * daniel_hozac base.__cleanup needs to happen last.
1160218122 M * daniel_hozac it removes the directories.
1160218127 M * mugwump aha
1160218144 M * mugwump so why does it work some times and not others.  intriguing
1160218166 M * mugwump ok, -a prompted me and left me with a shell.  don't forget to change the docs ;
1160218168 M * mugwump ;) even
1160218175 M * daniel_hozac are there any for -a?
1160218184 M * daniel_hozac -a seemed to be undocumented :)
1160218216 M * daniel_hozac (like most other options the build methods have)
1160218274 M * mugwump maybe renamed to --debug or something too
1160218329 M * daniel_hozac well, IMHO we have too many --debug switches already meaning set -x
1160218337 M * mugwump w.t.f. is remounting my /usr RO
1160218383 M * mugwump no kernel messages about emergency remounting
1160218383 M * daniel_hozac have you tried vnamespace -n bash, and then running secure-mount with similar parameters as the build script, and checking which (if any) causes it?
1160218421 M * mugwump can mount operations inside a namespace affect mounts on the outside?
1160218431 M * daniel_hozac they really shouldn't...
1160218432 M * mugwump eg, a remount?
1160218440 M * mugwump worth testing
1160218453 M * daniel_hozac but the entire process should run in a new namespace, so that seems to be what's happening.
1160218492 M * mugwump I think the answer to my question is.  Why yes, remounts do affect all instances of that mount :)
1160218520 P * Viper0482 und weg
1160218520 Q * mnemoc Ping timeout: 480 seconds
1160218857 J * mnemoc ~amery@kilo105.server4you.de
1160219086 M * daniel_hozac hmm, and i guess HEAD would show other problems as well, as all the filesystems would be mounted nodev.
1160219112 M * daniel_hozac which filesystems are going to contain device nodes (that will be used)?
1160219152 M * mugwump the /usr/lib/fai/nfsroot has some dev nodes
1160219197 M * mugwump normally FAI extracts these from a tarball, however iirc my changes exclude /dev from being extracted from the base.tgz it makes
1160219237 M * daniel_hozac yeah, creating device nodes isn't allowed anyway. (no CAP_MKNOD in the vattribute)
1160219457 M * mugwump cool.  hey, looks like it's faireboot that's remounting stuff RO
1160219477 M * mugwump which happens if you interrupt the build.  I can probably fix that.
1160219500 Q * mire_ Quit: Leaving
1160219500 M * mugwump but it's an interesting caveat - secure_mount isn't so secure if you can remount
1160219531 M * mugwump unless the remount were to actually make a new mount then replace it in the current namespace
1160219545 A * mugwump hmms
1160219586 M * mugwump well, perhaps that is not a critical bug
1160219803 M * mugwump not from your perspective, anyway
1160219806 M * daniel_hozac and remounts affect all instances, not just the bind mount?
1160219819 M * mugwump I'll test that
1160219834 M * h01ger micah, i dont get your code: first you do VSERVERS=$vservers and then you operate with: if [-z $VSERVERS ] else for vserver in $vservers... is there any purpose in this other then confusion? ;-) 
1160219855 M * mugwump yes, it does
1160219872 M * mugwump on vs2.0.1, anyway
1160219881 M * daniel_hozac aaah, so you don't have BME then?
1160219893 M * mugwump 2.0.1 has BME, no?
1160219902 M * daniel_hozac no, it was added in 2.0.2 for the stable branch.
1160220374 M * daniel_hozac so the nfsroot is the only mount that needs devices?
1160220399 M * mugwump yep
1160220412 M * daniel_hozac there is no chroot'ing to /tmp/target for post-install scripts or suc?
1160220424 M * mugwump well, create the nodes first
1160220437 M * daniel_hozac they'll still give EACCES when accessed.
1160220447 M * mugwump oh, right
1160220452 M * mugwump yes, that happens a lot
1160220453 M * daniel_hozac (and they are created rather early in the process)
1160220478 M * daniel_hozac but FAI_DIR doesn't have any?
1160220482 M * mugwump no
1160220487 M * daniel_hozac ok then.
1160220943 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/uv/experimental/vserver-build.fai i did some more cleanup, should remove the runfile now too on failure.
1160220968 M * daniel_hozac (or success, for that matter)
1160220985 M * mugwump save_ctxinfo: open("/var/run/vservers/ns"): File exists
1160220990 M * mugwump still get that if that file exists
1160221014 M * mugwump oh, maybe that's cleaned up
1160221020 M * daniel_hozac it should be.
1160221047 M * daniel_hozac (you'll have to remove it manually now, but then it should be taken care of)
1160221143 M * mugwump hmm, after exiting the shell left by -a, the processes still aren't all killed
1160221158 M * mugwump without -a it seems to be fine
1160221171 M * mugwump the only one left is the sh -i, bizarrely
1160221193 M * daniel_hozac that's very strange...
1160221207 M * daniel_hozac so you get the warning message?
1160221238 J * Markus23 ~Markus@86.59.55.224
1160221282 M * mugwump which warning were you expecting I see?
1160221305 M * daniel_hozac the most recent script should output a warning if the context is still around after the vkills.
1160221395 M * mugwump ok, I can't reproduce that, and looking back, it looks like I was using the old version of the script that time
1160221428 M * mugwump yeah, I think that'll do for now.  Perhaps refer to a wiki page for current information about which FAI packages are required...
1160221446 M * daniel_hozac rather than the SVN URL?
1160221474 M * mugwump yeah; I can set up some apt sources and put better information there
1160221501 M * daniel_hozac sure, do you have the wiki page ready? i.e. what URL should i put in there? :)
1160221524 M * mugwump http://faiwiki.informatik.uni-koeln.de/index.php/Using_FAI_as_a_vserver_build_method  # I'll fix it RSN ;)
1160221591 M * mugwump I've got to scram...
1160221645 M * daniel_hozac ok, thanks for testing!
1160221666 M * mugwump no worries ... thanks for working on this!
1160223684 Q * bubulak Remote host closed the connection
1160223696 J * bubulak ~bubulak@whisky.pendo.sk
1160224096 J * mire ~mire@76-167-222-85.adsl.verat.net
1160225334 M * shedi has anyone encountered this problem with the bind9 server while reloading "bindsocket.c:3122: 22/Invalid argument"
1160225374 M * daniel_hozac no.
1160225551 M * daniel_hozac what does strace say?
1160225738 M * shedi how do I use strace? :)
1160225920 M * daniel_hozac strace -fF -o trace <command that fails>, post the trace file when it's done ;)
1160226235 M * daniel_hozac ATTN: unless someone has an objection (or i find some more bugs while testing it), my intention is to release util-vserver 0.30.211 around 20:00 CEST today. (in about 5 hours)
1160226298 N * Bertl_zZ Bertl
1160226304 M * Bertl morning folks!
1160226332 M * Bertl daniel_hozac: ah, so the bug hunt will start a quarter past eight :)
1160226342 M * daniel_hozac hehehe, probably.
1160226367 M * daniel_hozac morning Bertl.
1160226397 M * Bertl shedi: you might want to attach to the reloading bind (if that causes the errors)
1160226442 M * daniel_hozac i think the error is from connect(), probably from rndc.
1160226445 M * Bertl daniel_hozac: what do you think about the schedmon interface (and the history one too, as it is a copy)
1160226460 M * shedi shall I paste the results here or in a paste bin
1160226464 M * daniel_hozac i haven't gotten around to reading the patch yet...
1160226465 M * shedi a few lines
1160226466 M * daniel_hozac shedi: pastebin.
1160226479 M * daniel_hozac (unless it's less than 4 lines)
1160226504 M * shedi http://paste.linux-vserver.org/453
1160226562 M * daniel_hozac shedi: what's more interesting is the file named trace that should be in your cwd.
1160226642 M * shedi I've tried numerous ways to deal with this
1160226821 J * _are_ ~are@62.112.159.81
1160226898 M * daniel_hozac Bertl: shouldn't the cmd structures use uint32_t since do_read_history expects that?
1160226961 M * Bertl good point :)
1160226987 M * Bertl didn't even see that ...
1160227015 M * daniel_hozac other than that and doener's NR_CPUS comment, i think it looks fine.
1160227043 M * Bertl okay, I already addressed that one
1160227791 J * lilalinux ~plasma@h1-gw.of.net-lab.net
1160228043 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.18-vs2.1.1-rc37.diff
1160228071 M * Bertl will now start coding the transfer client ...
1160228379 M * shedi regarding  my rdnc/named problem this is what happens when I run vserver stop http://paste.linux-vserver.org/454
1160228396 M * daniel_hozac should we update the topic/wiki for -rc37?
1160228403 M * daniel_hozac shedi: what happened to the strace file?
1160228484 T * Bertl http://linux-vserver.org/ <- new and shiny | latest stable 2.02.1, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.1.1-rc37, stable+grsec 2.0.2.1 | util-vserver-0.30.210 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;)
1160228484 M * shedi daniel_hozac, the output from strace running /etc/init.d/bind reload was this http://paste.linux-vserver.org/453
1160228500 M * daniel_hozac shedi: right, but what about the file named trace?
1160228539 M * daniel_hozac where the actual tracing would be.
1160228544 M * mnemoc *g*
1160228611 M * shedi daniel_hozac, I'm sorry I don't really understand
1160228633 M * daniel_hozac shedi: when you ran that strace command, it should've created a file named trace.
1160228643 M * daniel_hozac (in your current working directory)
1160228648 M * shedi k
1160228651 M * daniel_hozac the contents of that file is what's interesting.
1160228686 P * Markus23 Left the chatroom.
1160228712 M * shedi k I see what you mean
1160228925 M * shedi http://paste.linux-vserver.org/455
1160229023 M * daniel_hozac shedi: hmm, what kernel is that and what interfaces have you assigned to your guest?
1160229051 M * shedi 2.6.17.13-vs2.0.2.1 eth0 
1160229097 M * shedi but I'm having the same problem on earlier versions 
1160229206 Q * _are_ Ping timeout: 480 seconds
1160229213 Q * mugwump Read error: Connection reset by peer
1160229425 J * _are_ ~are@62.112.159.81
1160229475 J * mugwump ~samv@watts.utsl.gen.nz
1160229648 M * Bertl doener: do you still plan to work on the dump interpretation, or should I code that into the dumper?
1160229703 M * trippeh Is there any news on the network namespaces stuff that flew past LKML a while ago?
1160230006 Q * _are_ Ping timeout: 480 seconds
1160230117 M * Bertl trippeh: not that I know of ...
1160231624 A * Bertl .o( coding to metallica feels good :)
1160232885 J * Piet hiddenserv@tor.noreply.org
1160232924 M * Bertl wb Piet!
1160234086 Q * phedny Ping timeout: 480 seconds
1160234251 Q * Zaki Ping timeout: 480 seconds
1160234282 J * phedny ~mark@volcano.p-bierman.nl
1160236304 M * h01ger micah, monitoring VM doesnt work with vserver_resources
1160236358 M * h01ger (and btw, vserver_resources sucks as a (script|variable) name: its ressources in german or rather french, and resource in english. WTF?! :-))
1160236369 M * h01ger same as address... but vice versa :)
1160237015 J * Zaki ~fff@212.118.110.114
1160238150 M * h01ger micah, fixed it, commit 232
1160238851 Q * Johnnie Read error: Connection reset by peer
1160238865 J * Johnnie ~jdlewis@216.75.24.218
1160239555 J * fznsky ~second@61-231-87-194.dynamic.hinet.net
1160239975 M * Bertl welcome fznsky!
1160240029 M * fznsky hello
1160240237 P * fznsky 
1160240781 M * Bertl after applying this fix: http://vserver.13thfloor.at/Experimental/delta-smon-fix01.diff it looks like the monitor is working
1160240908 M * micah h01ger: oops, yeah that rework i did of the script was just something quick, there is some more work to be done
1160240934 M * Bertl doener: ping?
1160241031 M * h01ger micah, svn up - i fixed the errors. 
1160244163 M * daniel_hozac ok, no objections...
1160244167 M * daniel_hozac proceeding with the release then.
1160244194 M * Bertl oh noooo! yeah fine with me :)
1160244243 M * daniel_hozac hmm.
1160244249 M * daniel_hozac svn: Can't create directory '/var/svn/util-vserver/db/transactions/2354-1.txn': Permission denied
1160244271 M * daniel_hozac do you have permissions to fix that?
1160244281 M * Bertl I guess so, let me check
1160244502 M * doener Bertl: pong
1160244511 M * doener sorry, forgot to set away mode
1160244890 M * Bertl np, right on time
1160244910 M * Bertl http://vserver.13thfloor.at/Experimental/TOOLS/vdump-0.02.tar.bz2
1160244927 M * daniel_hozac ah, nevermind, was a problem with Hollow's script (http vs. svn+ssh).
1160245018 M * Bertl doener: with the latest patches applied, it seems to be able to gather data from the monitor (haven't tested the history part yet)
1160245047 M * doener Bertl: hm, interesting new indentation style ;)
1160245054 M * Bertl doener: so if you still want to code on the interpretation of the data .. be my guest ...
1160245074 M * Bertl doener: it's tab=8 indent=4 in userspace
1160245096 M * Bertl (probably not cleaned up yet)
1160245161 M * daniel_hozac ensc: ping
1160245618 M * Bertl daniel_hozac: did you move all the daniel.hozac.com stuff over to people.linux-vserver.org, or do I have to look at both locations?
1160245637 M * daniel_hozac for what? all of my linux-vserver related stuff should be on people.
1160245648 M * Bertl for patches :)
1160245675 M * Bertl so, /p/k then :)
1160245681 M * daniel_hozac yep, for kernel patches :)
1160245739 M * daniel_hozac ensc, micah, phreak``, Bertl, doener, everyone: util-vserver 0.30.211 has been uploaded to http://ftp.linux-vserver.org/pub/utils/util-vserver/
1160245751 M * Bertl yay!
1160245810 M * Bertl topic?
1160245838 T * doener http://linux-vserver.org/ <- new and shiny | latest stable 2.02.1, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.1.1-rc37, stable+grsec 2.0.2.1 | util-vserver-0.30.211 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 
1160245854 M * daniel_hozac thanks, i couldn't get it on one line...
1160245866 M * daniel_hozac (and tabbing it didn't work either, for whatever reason)
1160245916 M * doener daniel_hozac: irssi? sometimes you need to type "/topic  <tab>" (two spaces)
1160245930 M * daniel_hozac aah, thanks!
1160246458 M * Bertl doener: sometimes?
1160246511 M * doener It worked with one space the first time I tried... then it failed in #kernelnewbies and since I discovered that it works with two spaces there, I always use 2
1160246538 M * Bertl same here, never succeeded with one though :)
1160246548 M * doener actually, it works with one in here right now
1160246561 M * Bertl fascinating ...
1160246562 M * doener (unless you tab-complete /topic as well)
1160246571 M * daniel_hozac hmm, indeed.
1160246601 M * Bertl nope, no go here ... but I guess it's some kind of feature :)
1160246641 M * doener probably... to avoid letting it become business as usual or sth.
1160247587 Q * lilalinux Remote host closed the connection
1160248001 M * Bertl daniel_hozac: do you have an updated version of the cow-feat04?
1160248001 J * _are_ ~are@62.112.159.81
1160248071 M * _are_ hi
1160248087 M * daniel_hozac Bertl: not yet... give me a sec.
1160248105 M * daniel_hozac hello _are_.
1160248168 M * Bertl daniel_hozac: np, I'll take a nap now .. will be back later
1160248182 M * daniel_hozac ok, sleep well!
1160248187 N * Bertl Bertl_zZ
1160248387 M * daniel_hozac Bertl_zZ: http://people.linux-vserver.org/~dhozac/p/k/delta-cow-feat05.diff
1160248803 M * micah daniel_hozac: ok! I'll try and get it in
1160248869 M * micah daniel_hozac: i'm going to have to try and figure out which patches that are in the package are no longer needed
1160248910 M * daniel_hozac micah: IIRC all but the vserver-build.functions_etc.dpatch.
1160248929 M * micah nice, you are making this easy :)
1160248932 M * daniel_hozac umm, i meant the opposite.
1160248945 M * daniel_hozac i.e. all but that one are merged :)
1160248968 M * daniel_hozac what is the reason for that patch, anyway?
1160248973 M * micah right, so remove all the patches, except that one
1160248984 M * micah I need to look at it again
1160249057 M * micah when I first got hold of the ability to update the package, the first thing I did was try and get rid of as many of those patches as possible and submit them upstream
1160249068 M * daniel_hozac hehe.
1160249078 M * micah one sec and I'll look
1160249494 M * micah i remember there being one or two that I was puzzled by
1160249598 M * micah daniel_hozac: I think that one broke the debian-vservertools package of ola's
1160249606 M * micah or rather, the lack of that patch
1160249637 A * micah removes 20 patches
1160250002 M * micah i'm updating the testme and testfs scripts as well
1160250076 M * mnemoc which are the last versions of the test* scripts?
1160250129 M * micah mnemoc: testfs.sh-0.14 and testme.sh-0.16
1160250140 M * micah daniel_hozac: do you have an opinion about this: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=383794
1160250217 M * mnemoc micah: thanks :)
1160251146 M * daniel_hozac mnemoc: you can see that at http://vserver.13thfloor.at/Stuff/SCRIPT/ ;)
1160251184 M * mnemoc daniel_hozac: i know.... but i assumed micah had the numbers in his L1 :)
1160251279 M * daniel_hozac micah: IMHO vserver-copy shouldn't be updated, it's legacy stuff. the rsync changes make sense, but i'd be very reluctant about adding the hosts change.
1160251293 M * mnemoc daniel_hozac: what about freshmeat's util-vserver ?
1160251297 M * micah mnemoc: you were right, it was a cache hit for me :)
1160251305 M * mnemoc micah: :)
1160251310 M * daniel_hozac 127.0.0.2 isn't rewritten, and ::1 is defined as _the_ loopback in IPv6, IIRC.
1160251336 M * daniel_hozac mnemoc: hmm, i guess ensc will have to update that.
1160251452 M * mnemoc daniel_hozac: it seems it's "public"... so the fm admins moderate the added releases
1160251523 M * daniel_hozac hmm, what do you think would need to be updated at freshmeat?
1160251551 M * mnemoc info about the last release?
1160251583 M * daniel_hozac it hasn't been updated for over 2 years? :)
1160251608 M * mnemoc :)
1160251734 M * daniel_hozac but yes, the links should be updated.
1160251946 Q * dna_ Quit: Verlassend
1160252066 M * micah yeah, and the savanna page needs updating too
1160252090 M * daniel_hozac does it? where?
1160252102 M * daniel_hozac i thought savannah was updated already.
1160252292 M * micah daniel_hozac: the "Latest News box"
1160252307 M * micah [Announcement] util-vserver 0.30 posted by ensc, Friday 07/02/2004 at 23:24 UTC - 0 replies
1160252311 M * micah really old :)
1160252315 M * daniel_hozac well, same there, hasn't been updated for two years.
1160252326 M * micah yeah
1160252329 M * daniel_hozac it's only stable releases there ;)
1160252355 M * mnemoc but 0.30.211 is labeled *stable* now, right?
1160253236 M * daniel_hozac hmm? it's still alpha util-vserver.
1160253317 M * micah daniel_hozac: its uploaded to debian :)
1160253409 M * phreak`` yup, I'm also preparing the ebuild :)
1160253645 M * mnemoc daniel_hozac: so considering 2.4 support is abandoned, what that "alpha" means?
1160254660 M * daniel_hozac what 2.4 support is abandoned?
1160254678 M * daniel_hozac and the "alpha" really just means that things might change incompatibly between releases.
1160254684 M * daniel_hozac (AFAIK)
1160254733 M * eyck mnemoc: AFAIK 2.4 is supported, very well I might add
1160254774 M * eyck not something you should consider for new installation though
1160254929 M * mnemoc by abandoned i mean: how long since the last time the 1.2 code was touched?
1160254940 M * daniel_hozac does it need to be touched?
1160254944 M * daniel_hozac it's a stable tree.
1160254945 M * eyck huh? 
1160254949 M * daniel_hozac it only gets bug fixes.
1160254974 M * mnemoc daniel_hozac: alpha is a bit more that inter-release, but ok... i'll retract my comment. sorry.
1160255001 M * daniel_hozac sure, the code really isn't alpha quality.
1160255027 M * eyck hear, hear
1160255037 M * daniel_hozac i mean, it's stable, but it's still the "alpha" tree.
1160255042 M * mnemoc =)
1160255093 M * matti Hi there.
1160255100 M * daniel_hozac hello matti
1160255101 A * matti just back from Cornwall :)
1160255187 M * matti Amazing place in England.
1160255211 M * daniel_hozac mnemoc: besides, 2.4 is supported by the alpha tree as well ;)
1160255291 M * matti Especially the two bio-spheres in place called "Project Eden". Very nice.
1160255292 M * matti :)
1160255385 M * mnemoc daniel_hozac: :)
1160255431 M * mnemoc daniel_hozac: considering 2.4 support is not abandoned, what about a rediff to a more "recient" 2.4 release? ,-)
1160255494 M * daniel_hozac mnemoc: is that required? eyck?
1160255614 M * mnemoc daniel_hozac: in fact it applies, but having patches to the current version of the kernel may "show" it still supported
1160255673 M * daniel_hozac but then someone would have to verify that the patches work for the new kernel.
1160255697 M * daniel_hozac i.e. that everything that should be virtualized/isolated is.
1160255713 M * daniel_hozac (i don't know anything about 2.4 to know how big of a task that is)
1160255791 M * mnemoc i only know it apply :)
1160256122 M * eyck daniel_hozac: rediff is required, but it works, I'm maintaining such rediffs 
1160256151 M * daniel_hozac i.e. it fails on things other than the Makefile? or just offsets.
1160256157 M * daniel_hozac s/.$/?/
1160256188 M * daniel_hozac eyck: are you publishing those anywhere?
1160256211 M * eyck hmm, yup, there used to be a link in old wiki/page, as 'bastard patchset' 
1160256283 M * eyck I haven't published latest tested rediffs, since interest seems to have dried up though (ie 2.4.33.3 rediff is only in my personal tree)
1160256294 M * mnemoc eyck: beside EXTRAVERSION, what is it failing to apply for you?
1160256312 M * eyck hm, few things, I don't recall now, trivial though
1160256398 M * daniel_hozac eyck: i thought you were the only one using 2.4 (and still updating) :)
1160256437 M * eyck daniel_hozac, oh really? and I thought everyone is using 2.4
1160256475 M * eyck mnemoac: might be because I apply few hotfixes before applying vs patch
1160256483 M * daniel_hozac it seems like most people who are still using 2.4 are running something like 2.4.2x, if not older.
1160256517 M * eyck mnemoc: but as of 2.4.33.3 most of those are no longer needed
1160256548 M * eyck daniel_hozac:  I hope not on anything internet-facing.. 2.4.2x is almost as riddled with holes as 2.6.xx
1160256556 M * daniel_hozac i agree.
1160256575 M * daniel_hozac but, well... people do stupid things all the time :)
1160256646 M * eyck hmm, debian is keeping it's 2.4.27 fairly well-patched up
1160256721 M * eyck wrong hyphenation, 'fairly well patched-up', I'm pretty tired, sorry
1160256796 M * daniel_hozac micah: btw, http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=349140 should be fixed in 0.30.211 (if it's the bug i think it is).
1160257258 M * micah daniel_hozac: great, I'll ask ola to test
1160257663 M * daniel_hozac is there a way to subscribe to a package's bugs in the Debian BTS?
1160257910 M * daniel_hozac ah, found it.
1160257929 M * daniel_hozac packages.qa.debian.org
1160258699 J * Piet_ hiddenserv@tor.noreply.org
1160259118 Q * Piet Ping timeout: 480 seconds
1160260962 J * _node node@130.85.149.17
1160261154 Q * bonbons Quit: Leaving
1160262412 Q * meandtheshell Quit: exit (0);
1160264555 J * azazel ~azazel@81-174-44-47.f5.ngi.it
1160264564 M * azazel hi all
1160264664 M * daniel_hozac hello
1160264698 M * azazel any thougth of gradually rewriting utils-vserver scripts into an higher level lang?
1160264736 M * daniel_hozac ... bash isn't high-level enough for you?
1160264974 Q * Piet_ Read error: Connection reset by peer
1160265045 J * Piet_ hiddenserv@tor.noreply.org
1160265096 M * daniel_hozac micah: urgency=low means 0.30.210 will be in etch? or we hope for the best and it might make it?
1160265132 M * azazel ...it depends... i'm just wondering if it really fits that job, and how a drop in replacement written in e.g. python looks like ...and what pros and cons it can have..
1160265202 M * daniel_hozac using anything other than bash would cause additional hard requirements.
1160265225 M * daniel_hozac (additional == things not likely to be present on the system already)
1160265241 M * azazel yep, of course....
1160265241 M * daniel_hozac and pulling in entire languages like Python or Perl, that would just be awful.
1160265263 M * daniel_hozac would make a minimal util-vserver host probably twice the size it is now.
1160265334 M * azazel doh...it's host size really an issue?
1160265370 M * azazel s/it's/is/
1160265377 M * daniel_hozac for minimal test systems, for instance, yes, very much so.
1160265378 M * FaUl azazel: depends on the environment, it may be
1160265414 M * daniel_hozac and personally seeing things pulling in python/perl for no reason other than "it's possible" kinda pisses me off.
1160265463 M * azazel yeah... here i  agree with you
1160265553 M * azazel mmmm ... busybox is able to run them?
1160265570 M * daniel_hozac why wouldn't it be? they should be compiled with dietlibc, and thus statically linked...
1160265588 M * daniel_hozac but you'll need bash.