1190764947 Q * Supaplex Quit: config rehash - bbiam
1190766342 J * edible edible@12-216-231-163.client.mchsi.com
1190770815 J * quasisane ~sanep@c-76-118-191-64.hsd1.nh.comcast.net
1190771657 Q * hparker Quit: Quit
1190772084 M * mugwump strange, 'vserver foo exec /var/tmp/filename' is returning 'vlogin: execvp(): Bad address'
1190772182 M * mugwump ah, it's a +x script without a #!
1190772510 J * hparker ~hparker@linux.homershut.net
1190776436 J * Loki|muh_ loki@satanix.de
1190776449 J * bzed_ ~bzed@devel.recluse.de
1190776462 Q * Loki|muh Read error: Connection reset by peer
1190776463 N * Loki|muh_ Loki|muh
1190776742 Q * bzed Ping timeout: 480 seconds
1190777842 Q * transacid Ping timeout: 480 seconds
1190778363 Q * Piet Quit: Piet
1190785022 Q * FireEgl Ping timeout: 480 seconds
1190785098 M * mugwump anyone know how to enter a vserver and write a wtmp entry inside?
1190785117 Q * quasisane Read error: Connection reset by peer
1190785152 M * mugwump er, utmp and wtmp
1190785194 M * mugwump hmm, sessreg perhaps
1190785208 J * jmcaricand ~user@d83-179-255-49.cust.tele2.fr
1190785222 J * quasisane ~sanep@c-76-118-191-64.hsd1.nh.comcast.net
1190785621 M * mugwump ah, vserver $vserver exec /bin/login -f $user
1190786613 J * FireEgl ~FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa
1190789823 M * mugwump !?!@ suexec doesn't chdir(~)
1190791984 J * julien-vserver ~jujux@lns-bzn-52-82-65-119-127.adsl.proxad.net
1190791989 M * julien-vserver hello
1190792319 M * julien-vserver daniel_hozac : Do you prefer I test  2.6.20.4-vs2.2.0-rc21 (or 2.6.19.7-vs2.2.0-rc21) or the latest version of the stable branch ? 
1190792327 M * julien-vserver that's for my problem of tty
1190792342 Q * mcp Ping timeout: 480 seconds
1190792371 M * julien-vserver because I can see other entries in the changelog for versions of 17 Jul 2007 in the stable branch
1190793493 N * Bertl_zZ Bertl_oO
1190793518 M * Bertl_oO julien-vserver: I'd test with the latest version of vs2.2.x
1190793550 M * Bertl_oO we do not make new releases just for doing a release, they are fixing bugs we found so far
1190793644 M * julien-vserver yes that's sure, but I think that was danial who said that the bug has been corrected in 2.6.20.4-vs2.2.0-rc21 2.6.19.7-vs2.2.0-rc21
1190793665 M * julien-vserver maybe he wanted I test one of this version
1190793729 M * julien-vserver do you have a preference for the kernel version ?
1190793775 M * Bertl_oO no, I guess he just wanted to point out that kernels before that have the issue he was referring to, and kernels after that do not :)
1190793796 M * Bertl_oO so 2.2.0.3 should be fine for you
1190793804 M * julien-vserver ok
1190793847 M * Bertl_oO should be back a little later ... cya
1190793847 M * julien-vserver I test with the latest kernel version
1190793851 M * julien-vserver ok, cya
1190794076 J * Pazzo ~ugelt@213.215.138.237
1190794151 Q * Pazzo 
1190795004 J * dna ~dna@27-196-dsl.kielnet.net
1190795391 J * yang yang@yang.sponsor.oftc.net
1190795559 J * Prens ~ss@85.107.45.170
1190796379 J * DavidS ~david@p57A49626.dip0.t-ipconnect.de
1190796486 Q * Prens 
1190797315 J * CWC CWC@89-215-37-177.2073053861.ddns-lan.pl.ekk.bg
1190797456 M * CWC hi
1190797465 M * DavidS re
1190797574 M * CWC cwc = crack windows corporations (c)
1190797598 M * CWC nice hobby
1190797619 M * CWC fuck the windows all
1190797857 J * JonB ~NoSuchUse@130.227.63.19
1190797929 J * transacid ~transacid@transacid.de
1190798044 J * coderanger ~coderange@kantrn.stu.rpi.edu
1190798303 Q * CWC Quit: Client exiting
1190798701 M * JonB hi guys
1190798775 M * JonB wrong
1190798778 M * JonB eeerrh
1190798815 M * JonB i have some java apps that eat all memory. A normal machine makes this hard to find out where the problem is
1190798819 M * ard is rebootmgr legacy or current?
1190798826 M * JonB because the machine becomes unreponsive
1190798869 M * JonB can i limit a vserver guest to some ammount of memory and then from the host inspect the processes inside the guest when the guest runs out of memory?
1190798895 M * ard JonB : http://linux-vserver.org/Memory_Limits
1190798898 J * cedric ~cedric@rny93-2-82-66-66-30.fbx.proxad.net
1190798992 J * meandtheshell ~markus@85.127.114.140
1190799055 M * JonB ard: thanks
1190799068 M * JonB ard: but can i then inspect the guest processes from the host?
1190799071 A * ard is also looking in that part
1190799080 M * JonB ard: does strace and similar work from the host?
1190799112 M * ard well, you can vps faux the thing, or enter context id 1 to inspect the process, but you have to be in the right context to be able to stracet the thing.
1190799136 M * ard Anyway, java vm's and most multi-threaded applications tend to be un-straceable
1190799142 M * JonB oh
1190799161 M * JonB bloddy developpers!
1190799164 M * ard :-)
1190799185 M * JonB "it's not my process that eats the memomry, it must be a hardware error"
1190799193 M * ard usually you get one of the threads in state T, while strace doesn't know it
1190799198 M * ard wut?
1190799231 M * ard if you have sun java 1.5 or 1.6 you can also ask the jvm what's wrong
1190799262 M * ard the jvm tends to limit itself to the amount of memory you say it may use...
1190799306 M * ard So if a java process runs out of system memory, the jvm is badly configured.
1190799333 M * ard If the jave program runs out of java memory, thos *!$!@# developers made a memory leak again
1190799548 M * JonB back
1190799555 M * JonB boss just came to talk about this problem
1190799621 M * JonB the developper said "I'm in doubt how java could eat all the
1190799621 M * JonB memory because the only process uses a lot of memory is limited with 1.3 Gb
1190799621 M * JonB RAM for heap. "
1190799646 M * JonB but anyway, i will give them a vserver gust
1190799660 M * JonB then they can play arround with having multiple java installations and configurations
1190799873 M * ard well, you can look at the java's memory usage...
1190799884 M * ard it's a single multithreaded process...
1190799898 M * ard so doing top and M will show you how much it uses...
1190799921 M * JonB yes thats what i used to show it eats more than 1,3GB
1190799934 M * ard is it a 32 bit jvm?
1190799938 M * JonB or that all processes they run in total does
1190799939 M * ard or 64?
1190799947 M * JonB i'm not entirely sure
1190799951 Q * meandtheshell Quit: Leaving.
1190800019 M * JonB why does it make a difference?
1190800102 M * JonB if i use the /etc/vservers/vserver-name/rlimits/rss.hard
1190800109 M * JonB what should i write inside it?
1190800113 M * JonB just the number?
1190800830 Q * dna Read error: Connection reset by peer
1190800850 J * dna ~dna@27-196-dsl.kielnet.net
1190801418 J * ktwilight ~ktwilight@125.220-66-87.adsl-static.isp.belgacom.be
1190801830 Q * ktwilight_ Ping timeout: 480 seconds
1190802279 J * Julius ~julius@p57B25BB3.dip.t-dialin.net
1190802377 Q * julien-vserver Ping timeout: 480 seconds
1190802689 J * julien-vserve1 ~jujux@lns-bzn-33-82-252-55-166.adsl.proxad.net
1190802722 Q * bzed_ Ping timeout: 480 seconds
1190802862 J * bzed ~bzed@devel.recluse.de
1190803069 M * ard JonB : just the number of 4k pages
1190803092 M * ard if the vserver exceeds that, you will see stuff like: VM: killing process sshd(9829:#4)
1190803099 M * ard in your kernel log :-)
1190803118 M * JonB thanks
1190803159 M * ard Hmmmm
1190803162 M * ard dpkg-checkbuilddeps: Unmet build dependencies: gawk autotools-dev dietlibc-dev vlan iptables autoconf (>= 2.52) libbeecrypt6-dev (>= 4.1.2-2) libtool automake1.9 e2fslibs-dev
1190803174 M * ard build dependencies on vlan? Why?
1190803247 M * JonB because you can  specify a vlan for a vserver
1190803259 M * ard that's oldscool :-)
1190803267 M * ard or legacy actually
1190803276 M * ard but as a build-depend ...
1190803300 M * ard as a co-maintainer of vlan I really haven't a clue what can be in that.
1190803312 M * ard the macvlan patch isn't even in experimental, so...
1190803454 M * ard macvlans will be interesting btw... You can have multiple l2 interfaces on a single network (each interface has it's own mac address)
1190803486 M * ard good voor l2 failovers such as vrrp, but also good voor vservers I guess :-)
1190803554 J * sharkjaw ~gab@158.36.44.106
1190803612 J * virtuoso ~s0t0na@ppp91-122-59-6.pppoe.avangard-dsl.ru
1190803614 Q * virtuoso_ Read error: Connection reset by peer
1190803615 M * JonB i suppose
1190803634 M * JonB ard: can i see the number of 4k pages or calculate it some how?
1190803659 M * JonB ard: 4GB ram / 1024 / 1024 /4 = number of posible 4k pages?
1190803750 M * JonB maybe it's 4*1024*1024/4 ?
1190803812 M * ard you mean how much a vserver uses?
1190803817 M * ard vserver-stat shows something
1190803837 M * JonB no
1190803846 M * JonB ard: i have created a rss.hard file
1190803855 M * JonB the contents of that file should be a number.
1190803859 M * ard yes
1190803863 M * JonB My physical machine has 4GB ram
1190803894 M * JonB so i suppose the maximum number i can put in the file would be 4*1024*1024/4 to get 4kpages ?
1190803905 M * ard yes
1190803910 M * ard or slightly less
1190803942 M * ard you can put more into it, since rss.hard is about the amount of swap and memory in use (if I am correct)
1190804461 M * JonB true, but...
1190804469 M * JonB if i put in too much
1190804473 M * JonB and it alocates it all
1190804476 M * JonB the machine will swap
1190804482 M * JonB which will be unpleasent
1190804558 M * JonB so 996148 seems like 3,8GB ram
1190804696 N * Bertl_oO Bertl
1190804704 M * JonB hey Bertl 
1190804708 M * Bertl hey folks!
1190804787 M * JonB Bertl: if i have a vserver guest that runs out of memory. Can i inspect the processes inside the guest from the host?
1190804814 M * Bertl sure, from the spectator context you can inspect all of them
1190804821 M * fb_ JonB: vtop may be your friend
1190804871 M * JonB Bertl: what information can i actually get ?
1190804913 M * Bertl probably everything you can get otherwise too
1190804962 M * JonB Bertl: strace? ptrace?
1190804976 M * daniel_hozac ard: rebootmgr is very much legacy.
1190804988 M * Bertl JonB: well, that's not really information, that is interaction
1190805008 M * Bertl JonB: yo can get the whole /proc stuff for each process/thread
1190805025 M * Bertl *you
1190805079 M * JonB Bertl: but i actually have to enter the guest and run ls/cat to see the contents of guest /proc right?
1190805082 M * julien-vserve1 does vserver depend on smp ?
1190805093 M * julien-vserve1 did it ?
1190805101 M * Bertl JonB: no, you do that from the spectator context (xid=1)
1190805108 M * JonB Bertl: okay
1190805110 M * Bertl julien-vserve1: no
1190805161 M * JonB Bertl: seems like i can use vserver to help the developpers debug their ram eating program :-D
1190805166 M * julien-vserve1 ok
1190805840 M * jmcaricand I can't install a perl module from CPAN into a guest. Any idea ? 
1190805873 M * daniel_hozac we'd need details, like, why it can't be installed to have any kind of clue.
1190806006 Q * Julius Ping timeout: 480 seconds
1190806014 M * jmcaricand last line : "make had returned bad status, install seems impossible". If I use perl Makefile.PL ..... It's works
1190806076 M * jmcaricand With strace i have a lot of : ioctl(3, SNDCTL_TMR_TIMEBASE or TCGETS, 0xbf8f8f18) = -1 ENOTTY (Inappropriate ioctl for device)
1190806089 M * daniel_hozac that's all the output you get?
1190806163 M * jmcaricand Checking if your kit is complete...
1190806166 M * jmcaricand Looks good
1190806171 M * jmcaricand Writing Makefile for Apache::Dispatch
1190806176 M * jmcaricand     -- NOT OK
1190806215 M * jmcaricand If I install on host, It's works...
1190806362 M * jmcaricand daniel_hozac: and for you, cpan APerlModule works in a guest ?
1190806436 M * daniel_hozac last time i tried.
1190806638 J * Julius ~julius@p57B25BB3.dip.t-dialin.net
1190806989 Q * rorem- Remote host closed the connection
1190807349 J * rorem- ~roremtank@bzq-219-46-202.isdn.bezeqint.net
1190807377 M * ard Hmmm.... for i in test*;do vserver $i start & done
1190807384 M * ard did not work out as I planned it would
1190807399 A * ard has 40 vservers hanging in startup
1190807487 M * ard daniel_hozac : /me has just found the contrib/manifest.dat.pathsubst file which classifies it very well :-)
1190807564 M * DavidS ard: i believe you have to repeatedly "fg" those processes
1190807573 M * ard :-)
1190807601 M * Bertl ard: worked perfectly fine for me last time I did that
1190807805 M * ard http://paste.linux-vserver.org/6765
1190807807 M * ard :-)
1190808327 M * Bertl well, maybe your runlevel scripts are interactive
1190808336 M * ard :-(
1190808341 A * ard does a fg
1190808353 M * ard but I think that doesn't help anymore
1190808364 M * Bertl for the next time, check out the vserver runlevel scripts
1190808376 M * ard root      8274    15 test15         0.0  0.0   2464  1168 pts/44   T    13:25   0:00              |           |           \_ /bin/sh /etc/init.d/rc 3
1190808376 M * ard root      8318    15 test15         0.0  0.0   1584   376 pts/44   T    13:25   0:00              |           |               \_ stty onlcr
1190808377 M * Bertl (which handle that properly, including background start)
1190808396 A * ard sighs
1190808544 M * ard root@c32791:/etc/vservers# for ctx in test*;do chcontext --xid $ctx killall -CONT -u root;done
1190808673 J * julius_ ~julius@p57B2539D.dip.t-dialin.net
1190809076 Q * Julius Ping timeout: 480 seconds
1190809565 M * JonB how do i make vserver automatically find the next unused context number use it, +1 and store the new value for the next free context?
1190809645 Q * DavidS Ping timeout: 480 seconds
1190809650 J * DavidS ~david@vpn.uni-ak.ac.at
1190810082 M * julien-vserve1 Bertl: I tested with 2.6.21-vserver from backports.org and same problem, I compile a vanilla kernel with 2.2.0.3
1190810120 M * julien-vserve1 I didn't strace kernel from backports
1190810285 J * Piet ~piet@tor.noreply.org
1190810466 M * Bertl what version does the 2.6.21 from backports have? (Linux-VServer)
1190810499 Q * Aiken Quit: Leaving
1190810503 M * julien-vserve1 good question
1190810688 M * julien-vserve1 is there a simple way to see which version of vserver use a binary kernel ?
1190810744 M * DavidS with debian kernels it is easiest to check the changelog.Debian
1190810766 M * julien-vserve1 ok thanks
1190810855 M * julien-vserve1 ahah "Update vserver patch to 2.2.0" :/
1190810896 M * julien-vserve1 not sure that's exactly the 2.2.0 and not a rc or a pre release
1190811062 M * julien-vserve1 that's possible the entry in the changelog is from 10 Jul 2007
1190811240 M * JonB automatic context setup ?
1190811281 M * julien-vserve1 bash script and temporary file ?
1190811366 M * JonB julien-vserve1: so you are saying that the vserver build command does not increment /etc/vservers/.default/context.next?
1190811396 M * julien-vserve1 absolutly not, I didn't play with this feature
1190811413 M * julien-vserve1 that was a suggestion
1190811467 M * JonB okay
1190811536 J * Darkglow ~pdesnoyer@208.71.184.41
1190811550 M * Bertl wb Darkglow!
1190811555 M * Darkglow :-)
1190811560 M * Darkglow thanks
1190811587 M * Darkglow thanks to your  help, everything is running smoothly now ;-)
1190811596 M * Bertl excellent!
1190811644 M * JonB Bertl: is context.next automatically updated?
1190811701 M * Darkglow I have a couple questions regarding networking... first, I found out that I cannot do a tcpdump inside a vserver and that I would have to wait for ngnet or something for that to work. Any way to enable tcpdump without too much security risks ? and when is this ngnet coming ? ;-)
1190811749 M * Bertl JonB: I'd assume so, please ask daniel_hozac for details
1190811792 M * Bertl Darkglow: tcpdump is putting the entire interface in promisc mode and captures all packets
1190811810 M * Bertl Darkglow: if you really want to enable it, you can, but it isn't really advised
1190811815 M * JonB daniel_hozac: is the number in context.next automatically updated?
1190811859 Q * julius_ Remote host closed the connection
1190811863 M * Bertl Darkglow: mainline is working on completely virtualized network stacks, which add significant overhead, but allow you to do tthings like that in a (moderately) secure way
1190811888 M * Bertl Darkglow: once those spaces will be available, Linux-VServer will of course support them
1190811897 J * Julius ~julius@p57B2539D.dip.t-dialin.net
1190811942 Q * ruskie Read error: Connection reset by peer
1190812013 Q * sharkjaw Quit: Leaving
1190812059 M * Darkglow ok, not a big deal. I can tcpdump on the host. thanks for the answers.
1190812070 M * Bertl you're welcome! 
1190812098 J * ruskie ruskie@ruskie.user.oftc.net
1190812676 M * matti Bertl: :)
1190812706 M * Bertl matti: :)
1190812745 M * JonB what happens if 2 guests try to use the same context number?
1190812759 M * Bertl they will end up in the same context
1190812825 Q * Julius Remote host closed the connection
1190813161 M * JonB okay
1190813177 M * JonB apparently it looks like vserver .-.. build does not update context.next
1190813303 M * julien-vserve1 what does it mean concretely that two guest end up with the same context, they can be interact on each other ?
1190813341 M * mugwump they can't see each other's FS
1190813345 M * Bertl they would basically share the same namespace (which can be process space or network space or others)
1190813361 M * mugwump ...unless they have the same FS namespace
1190813395 A * mugwump stops trolling ;) and sleeps &
1190813632 M * JonB Bertl: is that bad?
1190813706 M * Bertl well, depends, they will not be separated :)
1190813724 M * Bertl it's like running apps on the same host ...
1190813729 M * Bertl is that bad?
1190813733 M * JonB yes
1190813741 M * Bertl then don't do it :)
1190813904 M * julien-vserve1 finished to config, compiling... in some hour we will know the truth
1190813949 M * julien-vserve1 can it be a problem that debian are configuring for smp but I don't have a smp machine ?
1190813959 M * JonB Bertl: there is no auto context option?
1190813993 M * Bertl what is auto context supposed to do?
1190814015 M * julien-vserve1 increment context number with vserver build
1190814038 M * Bertl AFAIK, the tools do exactly that by default, if you don't specify a context number
1190814043 M * JonB Bertl: no running vservers sharing the same context number
1190814056 M * JonB Bertl: but why does a start then complain
1190814088 M * Bertl JonB: as you are supposed to assign static context numbers for each guest (which are unique) the abovementioned cannot happen, no?
1190814120 M * JonB context: vc_net_create(): Invalid argument
1190814140 M * Bertl what is the context number?
1190814157 M * JonB Bertl: but i dont want to remember all this by myself, i want autoMAGIC
1190814166 J * Julius ~julius@p57B2539D.dip.t-dialin.net
1190814171 M * Bertl file a feature request to savannah
1190814172 M * JonB Bertl: there are no context file for this particular vserver
1190814180 M * Bertl which is a bug!
1190814206 J * Punkie ~Punkie@home.pekelny.net
1190814214 M * Bertl welcome Punkie!
1190814223 M * Punkie Hi Bertl ;)
1190814231 M * Bertl JonB: every guest has to have a context entry
1190814249 M * JonB Bertl: okay
1190814254 M * JonB i will try again then
1190814263 M * julien-vserve1 you can know the context number with vserver-stat
1190814266 M * JonB with having a context.next file
1190814268 M * Bertl JonB: do you use older tools?
1190814304 M * julien-vserve1 I think that old versions of utils-vserver doesn't create a context number file, or I didn't find them on debian stable (when I don't specify a context number with vserver build)
1190814317 M * JonB Bertl: 0.30.212-1
1190814333 M * Bertl so ancient tools, please upgrade
1190814343 M * JonB they are just 2 numbers apart
1190814416 M * Bertl it's fine for me, just don't complain about possible issues which have been fixed a long time ago :)
1190814439 M * JonB Bertl: sorry
1190814546 M * Punkie I have a problem, I am triing vs2.3.0.24 and when I mount partition with tagxid options, it tell Unrecognize mount options "tagxid", have I mistake anywhere or something changed from vs2.2.0? 
1190814597 M * Bertl the option was renamed 'tag' in the devel branch
1190814615 M * daniel_hozac JonB: of course.
1190814619 M * Punkie Thanks a lot :)
1190814660 M * Punkie It works ;)
1190814670 M * Bertl excellent! you're welcome!
1190814750 M * julien-vserve1 daniel_hozac: I know that's not for me your answer, but can you remember the question
1190814770 M * daniel_hozac hmm?
1190814782 M * julien-vserve1 daniel_hozac: JonB: of course.
1190814801 M * daniel_hozac that context.next is updated when used.
1190814826 M * julien-vserve1 oh yes, excuse me
1190815001 M * JonB where do i find a debian util-vserver newer than the one in etch, for etch ?
1190815009 M * daniel_hozac bp.o
1190815050 M * julien-vserve1 it doesn't include the last version
1190815056 M * daniel_hozac yet.
1190815058 M * Punkie I used 213 from backpports http://backports.org/debian/pool/main/u/util-vserver/
1190815059 M * julien-vserve1 only 0.30.213
1190815104 M * JonB i prefer getting the VERY latest so Bertl doesnt scold me *sob* ;-)
1190815113 M * daniel_hozac micah should be able to upload 0.30.214 to bp.o any day now.
1190815156 M * JonB what is bp.o ?
1190815162 M * daniel_hozac backports.org
1190815163 M * sid3windr back ports dot ork
1190815189 M * JonB okay
1190815197 M * JonB apt-get.org did not listen that
1190815216 M * daniel_hozac even packages.debian.org refers to it...
1190815249 M * julien-vserve1 i never found anything usefull on apt-get.org, but that's probably me
1190815267 M * sid3windr I did
1190815269 M * sid3windr 4 years ago
1190815270 M * JonB maybe i should search in all and not just stable
1190815277 M * sid3windr doesn't seem like it's been updated since 2 years or so though :p
1190815322 M * julien-vserve1 JonB: and sometimes that's incredibly simple to do a backport, just rebuild the source package from unstable with dependancies from stable
1190815419 M * JonB julien-vserve1: oh, never tried that, i always used checkinstall, and that doesnt work any more
1190815438 M * JonB julien-vserve1: how do i make it choose sources from unstable?
1190815499 M * julien-vserve1 I think I just added deb-src from unstable in sources.list
1190815514 M * julien-vserve1 not sure I'm clear
1190815557 M * julien-vserve1 deb-src http://ftp.us.debian.org/debian/ unstable main <- just this
1190815644 M * JonB julien-vserve1: and then apt-source install foobar?
1190815760 M * JonB nope, found out
1190815878 M * julien-vserve1 http://www.murrayc.com/blog/permalink/2006/04/21/building-modified-debian-packages/
1190815892 M * julien-vserve1 one of simplest way
1190815907 M * julien-vserve1 at least doesn't require pbuilder or so
1190815963 M * julien-vserve1 I think I launched directly "debian/rules binary" when the source package was untared
1190815996 M * julien-vserve1 but I don't think we can call it the debian way
1190816155 M * JonB thanks
1190816237 M * julien-vserve1 you're welcome
1190817047 Q * FireEgl Ping timeout: 480 seconds
1190817503 M * micah yeah, .214 migrated to Lenny just yesterday, so i've built a backport of dietlibc (0.31) and a backport of util-vserver and have just been testing them, I expect to upload them soon
1190817610 M * daniel_hozac cool.
1190817697 M * Bertl micah: excellent! thanks!
1190817752 M * micah i found one silly error, so am rebuilding, but after that I think it should be ready
1190817758 J * ema ~ema@rtfm.galliera.it
1190817772 M * JonB micah: what silly error was that?
1190817774 M * daniel_hozac you did include that patch i showed you, right?
1190817774 M * micah could have been sooner, but the ia64 buildd was stuck and I had to contact the maintainer to fix it
1190817779 M * micah daniel_hozac: yep
1190817781 M * JonB okay
1190817784 M * daniel_hozac great!
1190817824 M * JonB micah: do you know if that patch is also in the source from unstable?
1190817838 M * fb_ micah: is it possible and/or safe to build .214 against etch's dietlibc?
1190817853 M * micah http://buildd.debian.org/~jeroen/status/package.php?suite=unstable&p=util-vserver&a=ia64
1190817900 M * micah fb_: it is possible, but the 0.31 version of dietlibc from lenny has some fixes for sparc
1190817913 M * micah JonB: the patch is not included in the unstable version
1190817938 M * JonB micah: hmm, i suppose i need to replace my own build with yours once it gets into backports
1190818006 M * micah JonB: I can make it available immediately so you dont need to wait for backports processing
1190818050 M * JonB micah: that would be great, then i can stop my current build of unstable source
1190818093 M * daniel_hozac note that the patch only matters if you're using deprecated dynamic contexts.
1190818129 M * JonB how do i know if i am that?
1190818137 M * Bertl déjà vu
1190818157 M * daniel_hozac hehe, indeed.
1190818178 M * Bertl well, fortunately we now know that this is a glitch in the matrix :)
1190818245 M * fb_ daniel_hozac: does it make sense to upgrade to .214 with stable branch?
1190818333 M * daniel_hozac if you'd like to have the bugs it fixes fixed, or the new features, yes.
1190818823 J * julius_ ~julius@p57B2539D.dip.t-dialin.net
1190819284 M * micah JonB: http://people.debian.org/~micah/dietlibc and http://people.debian.org/~micah/util-vserver/0.30.214-3bpo1
1190819335 M * micah its been uploaded to backports, so it should be there relatively soon
1190819382 M * JonB micah: thanks
1190819481 J * meandtheshell ~markus@85.127.114.140
1190819567 M * Bertl wb meandtheshell!
1190819603 A * meandtheshell waves at #vserver :9
1190819809 M * julien-vserve1 maybe one day this kernel will be compiled
1190820034 Q * nospoonuser Quit: reboot router for kernel update
1190820115 Q * julius_ Ping timeout: 480 seconds
1190820120 Q * Julius Ping timeout: 480 seconds
1190820412 M * julien-vserve1 youhou can boot on :)
1190820457 M * julien-vserve1 hummm
1190820492 M * julien-vserve1 Bertl: do you think I really need latest version of vserver-utils for my tty problem ?
1190820739 M * Bertl no
1190820763 M * Bertl but it won't hurt to try it
1190821259 Q * meandtheshell Quit: Leaving.
1190821300 M * ard micah : do you know there is vlan as a build-depend?
1190821347 A * ard just recompiled your packages for etch...
1190821374 M * matti B as Bertl 
1190821376 M * matti :)
1190821410 A * matti is wondering how to apply sesame street logic here :)
1190821494 M * micah ard: yes, I know vlan is a build-depend... but why do you ask?
1190821555 N * ensc Guest64
1190821564 J * ensc ~irc-ensc@p54B4DA93.dip.t-dialin.net
1190821670 Q * Guest64 Ping timeout: 480 seconds
1190821679 M * Bertl okay, off for dinner ... back later
1190821685 N * Bertl Bertl_oO
1190822137 M * daniel_hozac ard: as it should be...
1190822264 M * julien-vserve1 argh forgot to include option --initrd :'(, have time to finish my battle for wesnoth 
1190822725 M * ard micah : I was just wondering why is it a build depend?
1190822772 A * ard can imagine a normal depend
1190822847 A * ard sees a test for the existence of vconfig in configure
1190823128 A * ard wonders if VCONFIG=/sbin/vconfig in the debian/rules is not enough?
1190823156 J * bonbons ~bonbons@2001:960:7ab:0:20b:5dff:fec7:6b33
1190823624 J * _jmcaricand_zzz ~jmcarican@d90-144-122-137.cust.tele2.fr
1190823830 M * micah ard: if you find out, let me know, and I will be happy to remove it and set that instead
1190823850 Q * JonB Ping timeout: 480 seconds
1190824198 M * ard micah : ok :-). If I have time I also wanted to split up according to contrib/manifest.dat.pathsubst
1190824449 A * ard is removing vlan, iptables, wget, procps, module-init-tools, net-tools from the environment and build-depends.
1190824570 M * daniel_hozac that sure seems future-proof *cough*
1190824575 M * ard 8-D
1190825231 A * ard understands
1190825254 M * ard at least procfs is required, so should always be there on a build system :-)
1190825846 A * ard sees a future for mvconfig in vserver.functions
1190827324 Q * ema Quit: leaving
1190827399 J * renihs ~narf@86.59.52.212
1190827831 M * igraltist hi
1190827865 M * igraltist soon i have must vguest ready compiled for using the X
1190827877 M * igraltist what i need to use the X in the vguestß
1190827878 M * igraltist ?
1190827909 M * igraltist must in vhost the Xserver also running?
1190827921 M * ard micah : http://paste.linux-vserver.org/6769
1190828292 J * meandtheshel1 ~markus@85.127.114.140
1190828414 J * besonen_mobile_ ~besonen_m@71-220-225-14.eugn.qwest.net
1190828477 Q * Piet Ping timeout: 480 seconds
1190828512 M * micah ard: if you set those variables and do not have vlan installed, you can still build?
1190828556 Q * transacid Ping timeout: 480 seconds
1190828743 M * ard yes
1190828758 M * ard those variables are just plain overrides of the configure test system
1190828766 Q * besonen_mobile Ping timeout: 480 seconds
1190828773 M * ard but daniel_hozac is right if he says : [18:36] <daniel_hozac> that sure seems future-proof *cough*
1190828774 M * ard :-)
1190828804 M * ard but for vlan I guess it won't ever be a problem.
1190828859 M * ard what is more interesting is to split util-vserver into a util-vserver and a legacy add on according to: contrib/manifest.dat.pathsubst
1190829076 J * Piet ~piet@tor.noreply.org
1190829129 J * FireEgl FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa
1190829425 J * coderanger_ ~coderange@marvin-44.dynamic2.rpi.edu
1190829514 J * Pazzo ~ugelt@213.215.138.237
1190829614 J * Julius ~julius@p57B2539D.dip.t-dialin.net
1190829669 Q * meandtheshel1 Quit: Leaving.
1190829678 A * ard is going home
1190829681 M * ard o/~
1190831436 M * julien-vserve1 igraltist: xserver only in guest, and try startx to see what you need, only devices i think (i didn't need CAP_SYS_RAWIO)
1190831569 Q * DavidS Quit: Leaving.
1190832575 J * ema ~ema@rtfm.galliera.it
1190832707 Q * fb_ Quit: hopefully quick reboot
1190832879 J * JonB ~NoSuchUse@kg0-231.kollegiegaarden.dk
1190833112 M * julien-vserve1 Bertl_oO: same problem with 2.6.22 + 2.2.0.3
1190833505 M * julien-vserve1 and there's no dbg version of xserver in etch...
1190833676 M * julien-vserve1 what can I do sir ? :)
1190833738 M * julien-vserve1 I will do a .trace as last time but I don't think we will learn something new
1190833771 M * daniel_hozac do you have the rt_sigaction line anywhere?
1190833862 M * julien-vserve1 I didn't do the trace yet
1190834014 M * julien-vserve1 i'm trying to login with ssh and there's no route to host... I'm cursed
1190834031 Q * duckx Read error: Connection reset by peer
1190834088 J * duckx ~Duck@tox.dyndns.org
1190834453 J * CWC ~CWC@89-215-37-177.2073053861.ddns-lan.pl.ekk.bg
1190834514 Q * duckx Quit: Client exiting
1190834567 N * phedny Guest83
1190834574 J * phedny ~mark@ip56538143.direct-adsl.nl
1190834890 M * julien-vserve1 I didn't select my network card
1190834920 M * julien-vserve1 do I need to select vserver debugging code?
1190834944 Q * ema Quit: leaving
1190834972 Q * Guest83 Ping timeout: 480 seconds
1190834999 J * coderanger__ ~coderange@corenet-171.dynamic.rpi.edu
1190835031 J * fb fback@red.fback.net
1190835047 M * julien-vserve1 hummm don't really need the network, with the -p option of strace, ok I try to strace
1190835081 M * julien-vserve1 I'm tired, I need it
1190835102 M * julien-vserve1 so do I need vserver debugging code enabled too ?
1190835127 M * daniel_hozac yes.
1190835133 M * julien-vserve1 ok
1190835155 M * julien-vserve1 anything else important to check in kernel configuration ?
1190835166 M * julien-vserve1 for debugging I mean
1190835187 M * daniel_hozac CONFIG_DEBUG_INFO is always good.
1190835298 Q * coderanger_ Ping timeout: 480 seconds
1190835302 M * julien-vserve1 history tracing and scheduling monitor? and which values if I need them
1190835424 M * daniel_hozac history might be useful, the scheduling monitor isn't interesting for this problem.
1190835430 M * julien-vserve1 ok
1190835467 M * julien-vserve1 do I keep the default value (64) ?
1190835472 M * daniel_hozac sure.
1190835479 M * julien-vserve1 That's MB ?
1190835493 M * daniel_hozac no, entires.
1190835496 M * daniel_hozac +english
1190835541 M * julien-vserve1 integer ?
1190835578 Q * hparker Quit: *burp*.. It's broke
1190835610 M * julien-vserve1 it doesn't make sense for me that's an history size of 64 integer...
1190835634 M * daniel_hozac it's 64 history entries.
1190835642 M * julien-vserve1 oooh ok
1190835667 M * julien-vserve1 you think that's enough so ?
1190835698 M * julien-vserve1 I prefer to ask more this time because it take really a long time on machine to compile a kernel
1190835794 J * hparker ~hparker@linux.homershut.net
1190835865 M * julien-vserve1 i'm going to eat, read you later
1190836276 Q * cedric Quit: cedric
1190836569 Q * _jmcaricand_zzz Quit: KVIrc 3.2.4 Anomalies http://www.kvirc.net/
1190836769 Q * renihs Quit: Leaving
1190837405 Q * the-dude Quit: Is that a kangeroo?
1190837412 J * the-dude ~martijn@senturparks.xs4all.nl
1190837442 Q * CWC Ping timeout: 480 seconds
1190837618 J * duckx ~Duck@tox.dyndns.org
1190837632 Q * JonB Quit: This computer has gone to sleep
1190837877 Q * Pazzo Quit: ...
1190837944 J * Aiken ~james@ppp121-45-249-108.lns2.bne4.internode.on.net
1190838427 J * _jmcaricand_zzz ~jmcarican@d90-144-122-137.cust.tele2.fr
1190838433 M * Radiance Hiya all
1190838486 M * Radiance about the remap source ip option in the vserver kernel section; does this means i don't need to configure services which need to listen only "locally" anymore ?
1190838500 M * daniel_hozac you've never had to do that.
1190838509 M * Radiance (just making sure i have a private ip as primary)
1190838585 M * Radiance daniel_hozac, ok, say i have a new vserver, and only internet ip's, what if for example i want mysql to only be reachable locally (like 127.0.0.1) how would i do that best ?
1190838614 M * daniel_hozac use a UNIX socket.
1190838626 M * Radiance please humor me, say there is no unix socket
1190838630 M * daniel_hozac or, use 2.3.
1190838668 M * Radiance and if i don't use 2.3
1190838688 M * daniel_hozac then give the guest a private address.
1190838704 M * Radiance normally i'd simply assign it a private range ip to listen on, and change /etc/hosts to make sure localhost == private range ip
1190838728 M * Radiance ok, i thought for a second that this "remap source ip" option in the kernel would make this method obsolete
1190838752 M * daniel_hozac no, it just rewrites the source address in addition to the destination address.
1190838789 M * Radiance ah i see, can you give me an example ? Just to make sure i understand this correctly mate :)
1190838826 M * daniel_hozac connections to 127.0.0.1 inside a guest will be redirected to the guest's first IP address.
1190838847 M * daniel_hozac with that option disabled, the source address will still be 127.0.0.1.
1190838856 M * _jmcaricand_zzz Hi. If I want to limit my guest to 1gb of memory (ram+swap) I can use this : vlimit -c 105 --rss $((2*512*256)). I'm right ? 
1190838857 M * daniel_hozac with it enabled, the source address will also be the guest's first IP address.
1190838897 M * Radiance daniel_hozac, i see now, so i guess in addition to the method i described i should enable this option too ?
1190838932 M * _jmcaricand_zzz perhaps swap doesn't exist in a guest ?
1190838939 M * daniel_hozac it sort of depens on how broken your software is.
1190838941 M * daniel_hozac +d
1190838952 M * daniel_hozac _jmcaricand_zzz: it's not really accounted for, no
1190838969 M * daniel_hozac _jmcaricand_zzz: but rss.hard - rss.soft is what will be shown as swap.
1190838969 A * bzed just submitted the first vserve rplugin to http://muninexchange.projects.linpro.no :)
1190838970 M * Radiance daniel_hozac, well, lets assume it's just the average LAMP config
1190838982 M * Radiance daniel_hozac, what would you advise in general for such config ?
1190838991 M * daniel_hozac bzed: i'm pretty sure there are munin plugins around...
1190839027 M * daniel_hozac Radiance: whatever works for you...
1190839045 M * Radiance ok thanks mate 
1190839053 M * bzed daniel_hozac: not on the munin-exchange unfortunately
1190839119 M * bzed daniel_hozac: there're 2 on http://svn.debian.org/wsvn/pkg-vserver/people/holger/munin-plugins/plugins/?rev=0&sc=0 though, which are linked somewhere in the munin wiki
1190839135 M * daniel_hozac yep, that's the one.
1190839147 M * bzed I'l ask holger to upload them
1190839168 J * drsource 0x2E2989@89-215-37-177.2073053861.ddns-lan.pl.ekk.bg
1190839567 Q * dna Quit: Verlassend
1190839592 Q * drsource Quit: ß åùå âåðíóñü ;-) http://www.web-hack.ru
1190839792 M * julien-vserve1 third kernel compilation today for vserver, i hope you're conscious you're priviledged (I hope that the problem is in the vserver patch...)
1190839818 Q * coderanger__ Read error: Connection reset by peer
1190839828 J * coderanger_ ~coderange@corenet-171.dynamic.rpi.edu
1190839927 M * daniel_hozac kernel compiles are typically very fast after the first one...
1190839972 M * julien-vserve1 hummm I'm surely doing something wrong so
1190840032 M * julien-vserve1 make menuconfig && make-kpkg clean && fakeroot make-kpkg --initrd --version=custom_0.1 ?
1190840044 M * julien-vserve1 maybe the make-kpkg clean ...
1190840059 M * daniel_hozac i don't know what make-kpkg does, but that doesn't sound good.
1190840243 M * julien-vserve1 which distribution do you used ?
1190840264 M * daniel_hozac Fedora and CentOS.
1190840333 M * julien-vserve1 ok
1190840336 M * fb julien-vserve1: you don't need make-kpkg clean on every run
1190840407 M * fb just don't change --revision between runnings
1190840457 Q * bonbons Quit: Leaving
1190840492 M * julien-vserve1 ok, I will know for next time
1190840531 M * julien-vserve1 everything is good to not be considered as priviledged :p
1190840682 Q * Julius Remote host closed the connection
1190841035 M * _jmcaricand_zzz julien-vserve1: I make all my vserver's kernel with make-kpkg. It's a very good tool :-) What is your problem ?
1190841135 M * julien-vserve1 I didn't read the man before to follow one doc found with google and did a make-kpkg clean each time I changed something
1190841141 M * julien-vserve1 that's only me the problem
1190841363 Q * FireEgl Read error: Connection reset by peer
1190841437 J * Punkie_ ~Punkie@home.pekelny.net
1190841542 Q * nebuchad` Read error: Connection reset by peer
1190841542 Q * Punkie Read error: Connection reset by peer
1190841555 N * Punkie_ Punkie
1190841556 M * bzed julien-vserve1: also you want to use make-kpkg --rootcmd fakeroot instead of runnign everythign in the fakeroot. speeds things up
1190841587 J * ruskie_ ruskie@goatse.co.uk
1190841593 M * bzed julien-vserve1: btw, why do you try to compile a kernel at all? there are vserver kernels in debian
1190841603 M * daniel_hozac but they're old
1190841609 M * bzed julien-vserve1: even if you'd be using ubuntu, you should be able to sue the debian kernels
1190841610 M * julien-vserve1 I have a problem running xorg in a guest
1190841617 J * nebuchad` ~nebu@82.233.222.74
1190841617 M * julien-vserve1 i'm using etch
1190841640 M * bzed daniel_hozac: linux-image-2.6.22-2-vserver-amd64 does not sound old
1190841643 M * _jmcaricand_zzz daniel_hozac: Is it possible to use this syntax for vlimit :  vlimit -c <VEID> --rss 512m  or it's not possible ? perhaps in future ?
1190841657 M * bzed julien-vserve1: if you need a more recent kernel, check backports.org
1190841679 M * julien-vserve1 i tried latest from backports.org too
1190841683 M * bzed I think the 2.6.22 is uploaded there
1190841703 M * julien-vserve1 but bp.o don't include 2.6.22 for powerpc, only for 386 and one other arch don't remeber
1190841710 M * daniel_hozac _jmcaricand_zzz: it's possible. note that it's 512 megapages, not MB.
1190841748 M * julien-vserve1 I'm no maso, if you have a simplier solution I take it
1190841778 M * bzed julien-vserve1: the powerpc buildd has a pretty much broken storage, shoudl be fixed soon hopefully
1190841805 Q * ruskie Ping timeout: 504 seconds
1190841834 M * bzed julien-vserve1: so patch your kernel, then configure it (copy and old config, use make menuconfigure and so on, what ever you ened)
1190841863 N * ruskie_ ruskie
1190841867 M * bzed julien-vserve1: then make-kpkg --initrd --revision=1 --rootcmd fakeroot binary
1190841880 M * julien-vserve1 all that is done
1190841893 M * bzed julien-vserve1: that should build you the whole package with docs and headers. replace binary by kernel_image if you need the image only
1190841953 M * bzed julien-vserve1: but I guess daniel_hozac knows better then me why your X doesn;t work
1190841956 M * bzed ;)
1190841981 Q * _jmcaricand_zzz Quit: KVIrc 3.2.4 Anomalies http://www.kvirc.net/
1190842067 J * FireEgl FireEgl@4.0.0.0.1.0.0.0.c.d.4.8.0.c.5.0.1.0.0.2.ip6.arpa
1190842133 Q * ruskie Quit: Changing server...
1190842140 J * ruskie ruskie@ruskie.user.oftc.net
1190843812 Q * Punkie Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.7/2007091417]
1190843829 J * Punkie ~Punkie@home.pekelny.net
1190843991 M * julien-vserve1 is there a way, when you launch startx to not automatically switch to the X11 vt ?
1190844654 J * onox ~onox@kalfjeslab.demon.nl
1190844898 M * Bertl_oO night folks ...
1190844907 N * Bertl_oO Bertl_zZ
1190845271 N * nebuchad` nebuchadnezzar
1190846264 J * meandtheshell ~markus@85.127.114.140
1190846349 Q * coderanger_ Quit: coderanger_
1190846716 J * coderanger_ ~coderange@xf-92a-17.dynamic2.rpi.edu
1190847726 M * julien-vserve1 compilation done, i will continue tomorrow, good night
1190847741 Q * julien-vserve1 Quit: Leaving.
1190848870 J * dowdle ~dowdle@scott.coe.montana.edu
1190848960 M * dowdle Greetings... probably an FAQ but... on a CentOS 5 VPS, trying to connect via ssh... I get the following in the /var/log/secure: pam_loginuid(sshd:session): set_loginuid failed opening loginuid
1190849002 M * daniel_hozac comment loginuid from /etc/pam.d/*, or disable auditing in the kernel.
1190849011 M * dowdle There's only the pam package installed... I wonder if there is some additional pam package I need?
1190849056 M * dowdle daniel_hozac: Any other way besides editing pam.d files?
1190849066 M * daniel_hozac disable auditing in the kernel, as i said.
1190849093 Q * onox Quit: leaving
1190849094 M * dowdle daniel_hozac: I can do that by just turning off the auditd service on the host?
1190849118 M * daniel_hozac no, you do that when building the kernel.
1190849125 M * daniel_hozac CONFIG_AUDIT.
1190849135 M * dowdle daniel_hozac: Using your CentOS 5 kernel... didn't compile it.
1190849166 M * daniel_hozac so, /etc/pam.d it is then.
1190849189 M * daniel_hozac i've been meaning to fix this in the kernel, but i haven't gotten around to it yet.
1190849302 M * dowdle daniel_hozac: Ok, thanks.  Only had to edit the sshd file.
1190849315 M * dowdle But I'm sure the others would fix other things.
1190849315 M * daniel_hozac if you want cron to work, you'll want to edit that too.
1190849349 M * dowdle Done.
1190849383 M * daniel_hozac i already have a patch that i think should work, but the last time i tried it, it didn't.
1190849430 M * dowdle daniel_hozac: Let me know if I become irritating.  So, how do I differenciate between i386 and x86_64 when creating a CentOS 5 client?  So far I just manually edited the yum.repos.d/.repo file to include the arch.  Any smarter way to do it?
1190849445 M * daniel_hozac prepend the command with linux32...
1190849491 M * dowdle Thanks.
1190849509 M * dowdle I have a lot of FAQ text to read still.
1190849584 M * daniel_hozac just remember that you'll want to use linux32 for any subsequent vyum calls, and that you should set the personality in /etc/vservers/<guest>/personality
1190849666 M * dowdle daniel_hozac: Ok.  I don't think I'll do that at the moment as I installed yum within the VPS.
1190849691 M * dowdle I don't see a personality file.
1190849697 M * daniel_hozac yes, but in order to do so, you'll need to prepend the vyum command with linux32.
1190849706 M * dowdle I just echo "linux32" > personality
1190849707 M * daniel_hozac you have to create it.
1190849710 M * daniel_hozac yes.
1190849734 M * dowdle If I do create personality I don't need to prepend?
1190849778 M * daniel_hozac not for vserver ... exec, i don't think vyum uses the personality..
1190849798 M * dowdle Ok.  What does use personality?
1190849818 M * daniel_hozac vserver .. exec/enter/start/stop/etc.
1190849837 M * bzed hmm, I never needed to set the personality at all, my i386 guests work just well... what's the reason behind setting it!?
1190849899 M * dowdle My i386 VPS is running as well without it...  Maybe it thinks it is x86_64?
1190849909 M * bzed my host is amd64
1190849933 M * bzed I'm running i386 guests in there, for example to get a browser with a proper java plugin
1190849982 M * daniel_hozac copy an x86_64 binary in there.
1190849982 M * dowdle Hmm, seems kernel logger starts up ok but doesn't want to shutdown at stop.  Weird.
1190850009 M * daniel_hozac in a linux32 guest, executing it won't work. without the personality, it'll work just fine.
1190850041 M * daniel_hozac (IIRC, anyway. it's been quite some time since i looked at it)
1190850060 M * dowdle Hmmm, I see the difference.
1190850102 M * dowdle I installed all i386 packages... but without personality set to linux32... when I start it and enter... and do a uname it says it is x86_64.  With personality, uname claims i686
1190850128 M * dowdle I see the difference... so it did think it was x86_64 even though all of the packages were i386.
1190850244 M * dowdle Is this the easy way to set timezone? ln -s /usr/share/zoneinfo/MST7MDT /etc/localtime
1190850255 M * bzed dowdle: then that stuff is imho pretty much broken. programs should not rely on the output of uname
1190850275 M * daniel_hozac dowdle: yes.
1190850284 M * daniel_hozac bzed: and how else should they figure out what the arch is?
1190850380 M * bzed daniel_hozac: I doubt you install sparc installations on i386 hardware
1190850409 M * daniel_hozac i don't see what you're getting at.
1190850415 M * dowdle bzed: What is broken and who said programs should depend on uname output?  I was just using it for my own info.
1190850450 M * daniel_hozac x86_64 vs. i386 is an interesting distinction for applications.
1190850460 M * bzed dowdle: I thought your yum or however ti was called depended on linux32
1190850470 M * daniel_hozac it does.
1190850487 M * bzed daniel_hozac: that's why debian uses dpkg-architecture to figure out which architecture you're building on and which architecture you're building for
1190850499 M * daniel_hozac it has nothing to do with building.
1190850499 M * dowdle bzed: I have i386 hardcoded in my .repo files... so it pulls the right packages... or at least it has been working for me even though I hadn't set personality.
1190850502 M * daniel_hozac it's for installing.
1190850534 M * daniel_hozac and, does dpkg-architecture use telepathy?
1190850551 M * daniel_hozac surely it too gets the architecture from somewhere, uname being the most obvious choice.
1190850555 M * daniel_hozac (as a default)
1190850592 M * bzed isntalling is a different thing, I jst tell debootstrap or whatever I use which architecture I'd like to have installed
1190850606 M * daniel_hozac and if you don't?
1190850636 M * dowdle Doesn't hurt to check... but I forgot how to get the arch version from rpm -qa?!?
1190850660 M * daniel_hozac rpm -qa --qf '%{NAME}-%{VERSION}-%{RELEASE}.%{ARCH}\n'
1190850676 M * bzed daniel_hozac: dpkg-architecture uses the output of the default compiler
1190850682 M * dowdle Yeah, that's it.
1190850703 M * bzed daniel_hozac: look, you can install an i386 system and add an amd64 kernel later
1190850734 M * dowdle Ok, good... everything is either i386, i686, noarch or (none)... so no x86_64 packages in there by mistake.
1190850737 M * bzed daniel_hozac: uname -r gives x86_64, which is just wrong if you'd start to use that, as your whole system is different
1190850750 M * dowdle But I'll know about prepending linux32 in the future and about personality.
1190850780 M * bzed daniel_hozac: same works at least for sparc afaik
1190850796 M * daniel_hozac sparc is different. you don't really want a 64-bit userspace.
1190850803 Q * coderanger_ Quit: coderanger_
1190850808 M * bzed no, but you could run it with a 64bit userspace in theory
1190850836 M * dowdle What's the range of values that context can be set to?
1190850842 M * daniel_hozac 2-49151.
1190850879 M * dowdle I'll start pulling my head out eventually... as more of this stuff sinks in.
1190850883 M * daniel_hozac as for using x86_64, i'd say it's the right thing to do. if i install an x86_64 kernel and then some application, odds are i'd want the 64-bit version.
1190850937 M * bzed depends, people prefer i386 for their desktops
1190850953 M * dowdle I see all of this stuff in /etc/vserver/<guest>  So, does all of that stuff have to be in separate files?  No single config file with all the info?  Or is that kinda like a psydo proc or something?
1190850955 M * daniel_hozac "people"? i'd love to get x86_64 for my workstation.
1190850974 M * bzed daniel_hozac: you're a geek and probably don't surf youtube all day and night
1190850987 M * daniel_hozac that's why i have a 32-bit firefox...
1190850996 M * dowdle If flash worked on x86_64 people might start using it on desktops.  So, what'll happen when the i386 / x86_64 arch merge happens in the mainline Linux kernel?
1190851007 M * daniel_hozac _one_ application that suffers the pain that is x86, vs. my entire desktop.
1190851022 M * dowdle daniel_hozac: You have more than 4GB of RAM?
1190851028 M * bzed daniel_hozac: yeah, you know how to use a 32bit firefox. the dumb user not.
1190851030 M * daniel_hozac no, i have more than 1 GiB of RAM.
1190851039 M * hparker dowdle: I have flash on my desktops
1190851048 M * bzed dowdle: uhm what do you mean by 'merging in the mainline kernel' !?
1190851048 M * dowdle hparker: 64-bit flash?
1190851060 M * hparker dowdle: No, 32bit works fine
1190851066 M * daniel_hozac bzed: and the dumb user often changes their x86 kernel to an x86_64?
1190851096 M * dowdle bzed: The decision was made to merge the i386 and x86_64 arches into a single arch (like was done with flavors of PowerPC and S/390)... over the loud disagreement of the maintainer of it.
1190851098 M * bzed daniel_hozac: no, but they install i386 on their 64bit machines
1190851110 M * daniel_hozac bzed: i don't see how that's relevant at all.
1190851118 M * dowdle daniel_hozac: I don't have trouble seeing over 1GB of RAM with an i386 kernel.
1190851131 M * daniel_hozac dowdle: but there's a significant performance hit.
1190851165 M * dowdle daniel_hozac: People keep claiming that... and I've used PAE for upto 16GB of RAM just fine... but I really haven't seen anyone with benchmarks.
1190851167 M * daniel_hozac (with the default memory split)
1190851182 M * bzed you can use nspluginwrapper to run the flash plugin on 64bit firefoxes
1190851193 M * bzed kinds broken sometimes, though