1190851209 M * hparker I really prefer to just get better pr0n 1190851276 Q * Johnnie Ping timeout: 480 seconds 1190851278 M * dowdle Wow... time flies... must catch bus. I'll pester you guys later. 1190851281 P * dowdle Darn bus. 1190851390 J * coderanger_ ~coderange@tweety-73.dynamic2.rpi.edu 1190851852 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.pa.comcast.net 1190852546 Q * meandtheshell Quit: Leaving. 1190852559 Q * Piet Quit: Piet 1190852637 Q * FireEgl Ping timeout: 480 seconds 1190853162 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 1190853313 Q * hparker Quit: Quit 1190853440 J * hparker ~hparker@linux.homershut.net 1190853617 Q * Punkie Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.7/2007091417] 1190855605 Q * coderanger_ Quit: coderanger_ 1190855903 J * jescheng ~jescheng@proxy-sjc-1.cisco.com 1190856073 M * jescheng is it possible to add kernel module from within a vserver? 1190856731 J * coderanger_ ~coderange@tweety-73.dynamic2.rpi.edu 1190856925 J * JonB ~NoSuchUse@kg1-20.kollegiegaarden.dk 1190856970 Q * Hunger Read error: Connection reset by peer 1190857114 J * Hunger ~Hunger.hu@213.163.11.138 1190858264 Q * JonB Quit: This computer has gone to sleep 1190858476 Q * Johnnie Ping timeout: 480 seconds 1190858730 J * jmcarica` ~user@d83-179-190-66.cust.tele2.fr 1190858862 Q * jmcaricand Ping timeout: 480 seconds 1190859051 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.pa.comcast.net 1190859556 Q * Johnnie Ping timeout: 480 seconds 1190859714 Q * jmcarica` Remote host closed the connection 1190860137 J * Johnnie ~jdlewis@c-67-163-142-234.hsd1.pa.comcast.net 1190861864 Q * coderanger_ Quit: coderanger_ 1190861909 Q * zLinux Remote host closed the connection 1190869498 Q * opuk Quit: leaving 1190872802 Q * jescheng Remote host closed the connection 1190872817 J * jescheng ~jescheng@proxy-sjc-1.cisco.com 1190872853 J * sharkjaw ~gab@158.36.44.106 1190873222 J * JonB ~NoSuchUse@kg1-20.kollegiegaarden.dk 1190873841 M * hparker Trying to compile openvcp, seems to not be happy with vserver.h http://rafb.net/p/wisA7I74.html Any ideas? 1190873891 M * hparker util-vserver 0.30.214, kernel 2.6.22-vs2.2.0.3-gentoo-r2 1190873905 M * JonB i wonder what openvcp is? 1190873922 M * hparker openvcp.org 1190873926 M * JonB found it 1190873941 M * hparker Some control panel gizmo I thought i'd look at :P 1190874024 M * JonB i dont think i can find it here http://linux-vserver.org/Downloads 1190874047 M * hparker yeah 1190874073 M * hparker thought maybe someone here was using it 1190874557 J * jmcaricand ~user@d83-179-222-124.cust.tele2.fr 1190874891 M * jmcaricand good morning. 1190875188 M * JonB hi 1190875339 M * jmcaricand In france, the weather in is bad. It's raining a lot :-( 1190875390 M * JonB poor you 1190875469 Q * JonB Quit: This computer has gone to sleep 1190875526 J * dna ~dna@17-192-dsl.kielnet.net 1190876293 J * DavidS ~david@p57A4A9E1.dip0.t-ipconnect.de 1190878548 J * JonB ~NoSuchUse@kg0-231.kollegiegaarden.dk 1190878611 Q * coderanger Remote host closed the connection 1190878626 M * arachnist sa 1190878640 M * arachnist same here in poland 1190878654 M * ruskie :) 1190878661 M * ruskie just overcast around here :) 1190878673 M * arachnist :P 1190878954 J * coderanger ~coderange@kantrn.stu.rpi.edu 1190879202 J * dna_ ~dna@15-215-dsl.kielnet.net 1190879608 Q * dna Ping timeout: 480 seconds 1190879809 Q * FireEgl Read error: Connection reset by peer 1190880493 M * DavidS overcast and cold ... welcome to the endless september ;) 1190880603 M * ruskie :) 1190880661 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 1190880793 N * Bertl_zZ Bertl 1190880798 M * Bertl morning folks! 1190880802 M * JonB hey be 1190880803 M * JonB Bertl: 1190880821 M * DavidS morning Bertl! 1190880888 M * Bertl JonB: what's up? :) 1190881065 M * JonB Bertl: hparker said that openvcp.org did not like vserver.h 1190881141 M * JonB oh wait, you said earlier you had a scroll log 1190881391 M * Bertl first, that is from util-vserver, so daniel_hozac is the one who might know 1190881403 M * JonB ok 1190881407 M * Bertl second, it looks to me like some types are not defined? 1190881427 M * Bertl so maybe a simple include will do 1190881440 M * JonB Bertl: i was wondering... do you get paied to do vserver development? 1190881467 M * Bertl depends 1190881507 M * Bertl some folks, like OLPC, pay for specific development 1190881558 M * JonB okay 1190881713 M * JonB how many more development ideas do you have? 1190881728 M * Bertl plenty of :) 1190881734 M * JonB heh 1190882307 Q * sladen Ping timeout: 480 seconds 1190883285 Q * JonB Ping timeout: 480 seconds 1190883306 J * pmenier ~pmenier@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1190884113 Q * DavidS Ping timeout: 480 seconds 1190884191 J * DavidS ~david@vpn.uni-ak.ac.at 1190884398 J * eviljonny ~eviljonny@loki.eviljonnys.com 1190884497 J * Punkie ~Punkie@home.pekelny.net 1190884968 N * dna_ dna 1190887032 J * sladen paul@starsky.19inch.net 1190887625 J * cedric ~cedric@rny93-2-82-66-66-30.fbx.proxad.net 1190887811 J * otaku42 ~otaku42@torvalds.h4ckr.net 1190887816 M * otaku42 moin 1190887823 M * daniel_hozac hi. 1190887896 M * Bertl welcome eviljonny! pmenier! cedric! otaku42! :) 1190887901 M * otaku42 running "apt-get upgrade" in a vserver with debian etch yields: failed to chmod '/usr/lib/pt_chown.dpkg-tmp': Operation not permitted 1190887911 M * otaku42 any hint on how to proceed looking for the cause is welcome :) 1190887911 M * Bertl wb Punkie! 1190887926 M * daniel_hozac otaku42: it's a known problem with dpkg. 1190887936 M * daniel_hozac otaku42: 2.2+ kernels work around it. 1190887937 M * cedric hey Bertl 1190887943 M * Punkie hi Bertl ;) 1190887965 M * eviljonny Hi Bertl 1190888377 J * JonB ~NoSuchUse@130.227.63.19 1190888467 M * sid3windr daniel_hozac: I bet his kernel is 2.6 ;) 1190888482 M * otaku42 daniel_hozac: host is running etch, too, with debian's stock verser kernel. there is one kernel upgrade pending. can i expect that this upgrade would help here, any idea? 1190888497 M * daniel_hozac otaku42: if the kernel upgrade is from backports, it should. 1190888534 M * otaku42 daniel_hozac: no, it's no backports kernel. 1190888561 M * daniel_hozac then no. 1190888587 M * otaku42 daniel_hozac: is there any other way to work around the issue? one which does not involve switching to a backports.org kernel, i mean? 1190888614 M * daniel_hozac IIRC there are patches for both dpkg and vhashify on the mailing list to work around it. 1190888662 M * otaku42 daniel_hozac: so it's an issue related to the use of vhashify? 1190888695 M * daniel_hozac more like an issue with the interaction between dpkg and unified files. 1190888832 J * sbauz ~sbauz@c211-28-60-223.eburwd8.vic.optusnet.com.au 1190888886 M * fb otaku42: kernel with grsec? 1190888924 M * daniel_hozac no, it's a vanilla Debian kernel. 1190888936 M * otaku42 fb: nope, without grsec 1190888939 M * fb hm 1190888968 M * fb i had similar problem, but with grsec kernel -- it denied to overwrite file with s 1190888993 M * fb but maybe one can do the same with vanilla kernel + vserver patch? 1190889003 M * daniel_hozac this particular problem? yes. 1190889049 M * Bertl daniel_hozac: please remind me, what is the specific dpkg issue? 1190889064 M * daniel_hozac it tries to chmod setuid binaries before unlinking the files. 1190889077 M * daniel_hozac and without COW, that leads to an EPERM. 1190889095 M * Bertl ah, i.c. so that's fixed in 2.3.x 1190889107 M * daniel_hozac and 2.2. 1190889109 M * daniel_hozac and 2.1. 1190889110 M * daniel_hozac :) 1190889115 M * Bertl excellent, tx :) 1190889290 Q * bzed Ping timeout: 480 seconds 1190889316 J * bzed ~bzed@devel.recluse.de 1190889418 Q * JonB Read error: Connection reset by peer 1190889440 P * sbauz Kopete 0.12.5 : http://kopete.kde.org 1190889680 J * JonB ~NoSuchUse@130.227.63.19 1190889923 M * otaku42 daniel_hozac: since i'm about to mess with a remote production server, i just want to double-check: linux-image-2.6.22-2-vserver-686, version 2.6.22-4~bpo40+1, is the right kernel from backports.org to fix the described issue? 1190890002 M * daniel_hozac no idea, i'd guess so. 1190890009 M * daniel_hozac i'm not a Debian user :) 1190890017 M * daniel_hozac (or, well, not an active Debian user) 1190890062 M * jmcaricand otaku42: why backport ? You can build your kernel with make-kpkg. If I can, you cant too. 1190890123 M * jmcaricand s/cant/can/ 1190890151 M * otaku42 jmcaricand: installing a ready-made package is less hazzle than compiling myself. both options share the same amount of risk anyway. 1190890384 M * jmcaricand I compile my vserver kernels myself for a lot of production servers without problem. You can use 1190890391 M * jmcaricand luke:/home/jmcaricand# 1190890405 Q * virtuoso Ping timeout: 480 seconds 1190890508 M * fb otaku42: that's one thing i'm not using from debian, the kernel. And there's no problem to prepare initrd-ready kernel with make-kpkg 1190890580 M * jmcaricand with --intrd option, make-kpkg makes all job -:) 1190891233 M * eviljonny Is it possible for the guest servers to talk to the host via IP, perhaps using the loopback interface? When I try pinging 127.0.0.1 I get 'ping: sendmsg: Operation not permitted'. Im using debian with iproute2 and the main interfaces for each verser are vlans using real public IP's. FYI I am trying to run a database on the host server which all of the guest servers have access too. 1190891276 M * eviljonny I also tried pinging the real IP of the host and also get operation not permitted. 1190891311 M * daniel_hozac are you running ping as root on the guest? 1190891333 M * eviljonny Yes 1190891333 M * daniel_hozac what kernel are you running? 1190891343 M * eviljonny 2.6.17.13-vs2.0.2.1 1190891356 M * daniel_hozac do you have iptables rules on the host to limit the traffic between the guests/host? 1190891361 M * daniel_hozac hmm, why such an old kernel? 1190891381 M * eviljonny Because we are running the default packages from the debian sarge tree 1190891394 M * daniel_hozac that kernel isn't in sarge. 1190891402 M * eviljonny Isn't it! hmmm 1190891419 M * eviljonny I didn't build this server I've just been tasked with working on it, I will go and ask the guy who built it 1190891447 M * eviljonny Thanks 1190891753 J * dna_ ~dna@121-226-dsl.kielnet.net 1190891867 Q * sharkjaw Quit: Leaving 1190892113 Q * dna Ping timeout: 480 seconds 1190892193 M * otaku42 does anyone know of a recipe for grub for: reboot server with new kernel; if that kernel fails, server gets power-cycled and then starts with old default kernel; if new kernel boots ok, admin makes new kernel default. any hints? 1190892214 M * daniel_hozac it depends on grub. 1190892218 M * daniel_hozac +which 1190892236 M * otaku42 daniel_hozac: 0.97 1190892249 M * daniel_hozac with the grub in Fedora, i do echo "savedefault --default= --once" | grub; reboot 1190892387 M * JonB i dont believe the FAQ or documentation writes anything about backup of guests? if the guest resides on a filesystem mounted inside the host, then can i just take a backup of that mountpoint? Is there any files below that i should exclude? like proc? 1190892402 M * JonB i suppose i should take a backup of the /etc dir as well? 1190892526 M * otaku42 daniel_hozac: i guess that grub would complain load and clear if i enter the savedefault command inside the grub shell manually when it does not support that feature? 1190892837 M * daniel_hozac otaku42: i'd assume so. 1190892855 M * daniel_hozac JonB: that all depends on what you want to achieve :) 1190892859 M * otaku42 daniel_hozac: ok, thx... i'll be brave then and try :) 1190892871 M * daniel_hozac JonB: by default the mount points inside the guest, i.e. /proc, /tmp, etc. won't be visible on the host. 1190893156 M * JonB daniel_hozac: i want to be able to restore a machine 1190893212 M * JonB i might reinstall the host to get newer tools, but the guests i want easy restore 1190893595 M * ard JonB : if the vserver is down, your guest is nothing more than a directory tree which you can copy to another server or whatever :-) 1190893614 M * ard and some config in /etc/vserver// 1190893634 M * ard if you have those two, you have a fully working system 1190893646 M * ard (for moving, backup/restore etc...) 1190893888 Q * dna_ Ping timeout: 480 seconds 1190894242 M * JonB ard: i might not take it down 1190894447 M * JonB is it an error that my guests hosts file does not contain a localhost line? 1190894870 Q * Aiken Quit: Leaving 1190896851 J * coderanger_ ~coderange@xf-92a-17.dynamic2.rpi.edu 1190896917 M * otaku42 daniel_hozac: upgrading to the backports.org kernel helped to get rid of my problem. thanks a lot for the hint. 1190896930 M * daniel_hozac you're welcome! 1190896987 M * daniel_hozac JonB: if you want it to resolve, i guess so ;) 1190897616 Q * cedric Quit: cedric 1190897817 M * JonB daniel_hozac: i was wondering why it did not exist. Maybe because i used util-vserver 212 and not 214? 1190897900 J * Piet ~piet@tor.noreply.org 1190897912 M * daniel_hozac util-vserver doesn't really do anything to /etc/hosts, except for the debian initpost script. 1190897923 M * daniel_hozac you can put files in e.g. /etc/vservers/.defaults. 1190897982 M * JonB okay 1190898414 J * meandtheshell ~markus@85.127.109.29 1190898481 J * ktwilight_ ~ktwilight@125.220-66-87.adsl-static.isp.belgacom.be 1190898510 Q * ktwilight_ 1190898532 J * ktwilight_ ~ktwilight@125.220-66-87.adsl-static.isp.belgacom.be 1190898711 J * ema ~ema@rtfm.galliera.it 1190898717 Q * ktwilight Ping timeout: 480 seconds 1190899091 Q * ktwilight_ Ping timeout: 480 seconds 1190899568 Q * meandtheshell Quit: Leaving. 1190899596 J * ktwilight ~ktwilight@197.220-66-87.adsl-static.isp.belgacom.be 1190899631 Q * ema Remote host closed the connection 1190899881 J * meandtheshell ~markus@85.127.109.29 1190900191 J * ema ~ema@rtfm.galliera.it 1190900465 Q * ktwilight Remote host closed the connection 1190900467 J * ktwilight ~ktwilight@197.220-66-87.adsl-static.isp.belgacom.be 1190900648 M * igraltist hi 1190900656 M * Bertl hey 1190900663 M * igraltist when i try to start the vserver guest i get this 1190900664 M * igraltist http://paste.debian.net/38150 1190900679 M * igraltist its looks thats he is always dieing 1190900708 M * Bertl what util-vserver version and what kernel? 1190900718 M * daniel_hozac looks like a suicidal guest to me. 1190900768 M * Bertl which reboots right on startup, yeah ... 1190900860 M * igraltist it was running, but then i do a lot of compiling inside the guest and today it does not start 1190900883 M * igraltist mayby i must start again fresh 1190900899 M * Bertl nah, compiling should not mess up your guest 1190900916 M * Bertl maybe a --debug run could help? 1190900940 M * igraltist Installed versions: 0.30.214 1190900953 M * igraltist this is the util-vserver 1190900973 M * igraltist and a normal kernel 2.6.22 1190900990 M * daniel_hozac what guest OS? 1190901031 M * igraltist gentoo 1190901042 M * daniel_hozac baselayout? 1190901042 M * igraltist normal desktop version 1190901065 M * igraltist i took a stage 3 template for vserver 1190901093 M * daniel_hozac a recent one? 1190901099 M * daniel_hozac from Hollow's people page? 1190901128 M * igraltist i dont remeber exactly 1190901154 Q * Piet Remote host closed the connection 1190901198 J * lilalinux ~plasma@dslb-084-058-240-191.pools.arcor-ip.net 1190901341 J * Piet ~piet@tor.noreply.org 1190901364 J * julius_ ~julius@p57B266E4.dip.t-dialin.net 1190901396 M * julius_ einen wunderschönen guten tag 1190901409 M * Bertl and a good day 2u2! 1190901576 M * JonB funny, my mysql server binds to the public ip adress, though only the localhost one is listed in my.cnf? 1190901621 M * sid3windr there is no localhost inside a vserver 1190901631 M * sid3windr it gets automatically translated to the first vserver ip 1190901639 M * Bertl ahem, no 1190901659 M * Bertl localhost is a symbolic name usually associated with 127.0.0.1 1190901671 M * JonB i just installed debian mysqlserver 5.0? 1190901677 M * JonB bind-address = 127.0.0.1 1190901677 M * Bertl and yes, 127.0.0.1 gets usually mapped to the first ip 1190901694 M * Bertl (or on recent devel kernels to the lback address) 1190901697 M * JonB Bertl: i do not understand what you mean with mapped 1190901713 M * Bertl but! and that is important, localhost can refer to any IP 1190901739 M * Bertl and usually, on older guests, it is advised to assign the first ip to the localhost name in /etc/hosts 1190901767 M * Bertl JonB: mapped means that binding to 127.0.0.1 will actually result in a binding to the first ip 1190901793 M * Bertl and packets sent to 127.0.0.1, will get remapped to 1190901800 M * JonB Bertl: do i have to do something special to enable that? 1190901806 M * JonB tcp 0 0 192.168.123.247:3306 0.0.0.0:* LISTEN 24985/mysqld 1190901825 M * JonB i would prefer having it only accessible from inside the guest 1190901826 M * Bertl I assume that is netstat not lsof output, yes? 1190901834 M * JonB that is netstat -a -p -n 1190901847 M * Bertl good, then it is already bound to 192.168.123.247 port 3306 1190901850 M * baggins Bertl: little something for you: http://paste.linux-vserver.org/6779 1190901851 M * JonB i'm not entirely sure that my my.cnf is correct 1190901866 J * dna ~dna@156-240-dsl.kielnet.net 1190901870 M * JonB Bertl: but... i would prefer mysql only to be accessible from inside the guest 1190901882 M * Bertl baggins: 2.6.22.9? 1190901902 J * julius__ ~julius@p57B266E4.dip.t-dialin.net 1190901915 M * baggins Bertl: yes, vs2.3.0.24 1190901932 M * Bertl did it apply cleanly? 1190901944 P * otaku42 1190901969 M * baggins no, I had to remove two hunks already present and do two little tweaks 1190901986 Q * coderanger_ Quit: coderanger_ 1190901987 M * baggins all were in fs code AFAIR 1190901998 M * Bertl hmm, and how do you trigger this one? 1190902062 M * baggins unfortuately I have no idea :( 1190902071 M * Bertl seems like lpd is involved 1190902096 M * Bertl do you have the kernel's build tree at hand? 1190902118 Q * julius_ Ping timeout: 480 seconds 1190902128 J * julien-vserver ~jujux@lns-bzn-33-82-252-4-58.adsl.proxad.net 1190902129 M * baggins yes 1190902132 M * julien-vserver hello 1190902160 M * Bertl baggins: okay, let's see what 'addr2line -e vmlinux ffffffff8044066a' produces 1190902165 M * Bertl wb julien-vserver! 1190902184 M * baggins ??:0 1190902194 M * Bertl bad, no debug info in your kernel 1190902244 M * baggins net/ipv4/inet_connection_sock.c:42 1190902250 M * Bertl ah? 1190902260 M * baggins rebuilt net/ipv4 with -g :) 1190902287 M * baggins if (ipv6_only_sock(sk2)) 1190902301 M * Bertl hmm, I do not really trust that line 1190902309 M * JonB Bertl: so the conclussion is that if i want to have a private mysql server i either have to tell it to stop networking or firewall it from the host? 1190902319 M * Bertl baggins: ipv4_rcv_saddr_equal+0x1aa/0x280 says it is somewhere in the second half 1190902376 M * Bertl JonB: private as in not reachable from outside the guest? 1190902381 J * coderanger_ ~coderange@pepe-11.dynamic2.rpi.edu 1190902390 M * baggins the nearest point I found while analyzing the code is 1190902392 M * baggins if (v4_addr_match(nxa, addr, tmask)) 1190902421 M * baggins it looks like line 62 1190902446 M * JonB Bertl: yes 1190902457 M * Bertl JonB: I would use unix sockets in this case 1190902470 M * JonB Bertl: i'll try that 1190902477 M * baggins Bertl: as for the second half remember about inlining 1190902490 M * Bertl JonB: because regardless on what ip you bind it or limit the access to, the host will always be able to reach it 1190902544 M * Bertl baggins: true ... 1190902575 M * julien-vserver JonB: can't you bind mysql to a specific ip, and setup a dummy interface ? 1190902600 M * julien-vserver oh ok 1190902602 M * julius__ 1) yes 2) not recommended 1190902611 M * JonB julien-vserver: it came with a bind-address = 127.0.0.1 1190902624 M * JonB julien-vserver: but it was still accessible 1190902635 M * Bertl baggins: hum hum ... that looks interesting, a quick grep for nx_v4_addr_conflict() gives two different definitions :) 1190902642 M * julien-vserver yes, Bertl knows what he says , sorry 1190902676 M * julius__ is it possible to use multiple(many) unix sockets? 1190902709 M * julius__ I'd like to offer ssh access to some webspace customers 1190902720 M * julius__ who will be jailed of course 1190902779 M * baggins I'm starving, bbl 1190902788 M * JonB julien-vserver: personally i still find it a little anoying that a bind to 127.0.0.1 doesnt mean localhost only 1190902829 M * julien-vserver maybe do it mean physical localhost, so a host or a guest is the same localhost 1190902833 M * Bertl julius__: how is that related to unix sockets? 1190902839 M * julius__ well 1190902855 M * julius__ they cant use the default socket 1190902871 M * Bertl you are talking about network port 22, yes? 1190902887 M * julius__ no 1190902890 M * julius__ mysql 1190902894 M * julius__ sry 1190902899 M * JonB julien-vserver: i mean lo inside the guest is local to guest only 1190902915 M * julius__ lighttpd + spawn-fcgi + mysql + ssh 1190902936 M * Bertl julius__: and how is that all connected? 1190902959 M * julius__ just forget it :) 1190902964 M * Bertl okay :) 1190903000 M * julius__ I'm to tired to understand what i'm thinking 1190903005 M * julien-vserver ahahah 1190903030 M * Bertl julius__: k, get some sleep, we talk about that tomorrow then :) 1190903139 M * julius__ ^^ 1190903154 M * julius__ one question that should make sense: 1190903178 M * julius__ can it be problematic to define a loopback device in multiple vhosts? 1190903201 Q * lilalinux Remote host closed the connection 1190903245 M * Bertl unfortunately that one doesn't make much sense either ... at least not for me :/ 1190903297 M * julien-vserver JonB: I'm a bit confuse as it seems it depends on the version you are running how loopback interface in guest are managed (they are jailed or not) 1190903323 M * JonB julien-vserver: hmm hmm, actually i have no lo interface 1190903336 M * Bertl julien-vserver: we do not 'jail' any loopback interfaces :) 1190903349 M * JonB Bertl: why not? 1190903375 M * Bertl because they are nice and friendly so imprisoning them would be against our morale :) 1190903400 M * julien-vserver by "jail" I mean each guest can see only their lo interface 1190903437 M * Bertl as interfaces per se do not 'beong' to the guest (they always belong to the host) that is dubious 1190903463 M * Bertl but 2.3.x (recent versions) do provide improved 127.0.0.1 virtualization 1190903473 M * Bertl (by using the lback address to remap to) 1190903497 M * Bertl *belong 1190903552 M * julien-vserver ok 1190903859 M * julius__ so it's some kind of crime do define a loopback device (dev => "lo", ip =>"127.0.0.1") for a guest? 1190903901 M * Bertl yes, and that will be punished with immediate termination of secure guest behaviour 1190903924 M * julius__ fine 1190903932 M * Punkie :D 1190903945 M * julien-vserver ok I understand a bit more, guests are good, humans are bad 1190903987 M * Bertl julius__: note: guests should be perfectly fine with bindings to 127.0.0.1 without that 1190904041 M * Bertl julius__: on kernels < 2.3.x, the ip will be remapped to the first guest ip, on devel kernels, it will get remapped to the lback ip, which will 'look' inside the guest like 127.0.0.1 1190904182 Q * matti Ping timeout: 480 seconds 1190904272 M * tam I have ipv6 support on my host, how can I get ipv6 inside a guest? I found the oldwiki page but it appears to be outdated. 1190904388 M * Bertl tam: get a recent patch (vs2.2.x + ipv6 patches or vs2.3.x) 1190904401 M * Bertl tam: and ipv6 inside a guest will work like ipv4 1190904429 M * tam thank you! 1190904441 M * Bertl you're welcome! 1190904586 M * tam Where can I find the ipv6 patch? 1190904614 M * Bertl for what kernel and vserver patch? 1190904635 M * tam Kernel: 2.6.20.4-vs2.2.0-arbor-1 1190904635 M * tam VS-API: 0x00020200 1190904635 M * tam util-vserver: 0.30.213; Jul 27 2007, 10:19:17 1190904699 M * tam http://people.linux-vserver.org/~bonbons/ipv6/patch-2.6.20.4-vs2.2.0.ipv6-rc19.diff 1190904709 M * tam That one? When I first looked at that page it looked unmaintained 1190904736 M * tam I didn't scroll down far enough 1190904743 M * Bertl http://people.linux-vserver.org/~dhozac/p/k/patch-2.6.20.14-vs2.2.0-ipv6.diff 1190904769 M * Bertl I would suggest updating to 2.6.20.14 (or later) and using that patch on a final 2.2.0 1190904788 M * Bertl you'll also need to update util-vserver to 0.30.214 1190904798 M * tam ok, thank you again! 1190904806 M * Bertl np 1190905249 M * Bertl okay, time for a nap .. back later ... 1190905253 N * Bertl Bertl_zZ 1190905303 J * matti matti@acrux.romke.net 1190905685 Q * JonB Ping timeout: 480 seconds 1190906215 J * dowdle ~dowdle@scott.coe.montana.edu 1190906267 M * julien-vserver daniel_hozac: there's no rt_sigaction in trace with kernel 2.6.22 and vserver 2.2.0.3 on powerpc when I switch from tty7 (xorg in guest) to tty1 (terminal in host) 1190906384 M * julien-vserver and I'm a newbie in kernel debug, so I need instructions on how to proceed to give you more information 1190906402 M * jmcaricand hi dowdle. how are you ? 1190906446 M * dowdle jmcaricand: Oh, pretty good. 1190906759 M * baggins Bertl_zZ: I went through the code and I think you did an overkill, see http://paste.linux-vserver.org/6780 1190906860 M * baggins Bertl_zZ: the call to ipv6_only_sock was never there where you put ipv4_rcv_saddr_equal 1190907405 J * zLinux ~zLinux@88.213.33.208 1190907419 Q * zLinux Remote host closed the connection 1190907775 J * zLinux ~zLinux@88.213.33.208 1190907780 M * daniel_hozac baggins: yes it was. 1190908038 Q * ensc Ping timeout: 480 seconds 1190908109 Q * coderanger_ Quit: coderanger_ 1190908146 J * ensc ~irc-ensc@p54B4CEB8.dip.t-dialin.net 1190908829 M * julien-vserver daniel_hozac : help please 1190908845 M * julien-vserver or say it if you don't have time now 1190908867 M * daniel_hozac echo 128 > /proc/sys/vserver/debug_misc and redo it. 1190908876 M * julien-vserver thanks 1190909378 Q * Piet Quit: Piet 1190909384 J * JonB ~NoSuchUse@192.38.8.25 1190910244 M * baggins daniel_hozac: sorry, you're right it was, but it wasn't where you substituted simple == with a call to ipv6_only_sock 1190910255 M * baggins *ipv4_rcv_saddr_equal 1190910303 J * coderanger_ ~coderange@xf-92a-17.dynamic2.rpi.edu 1190910360 M * baggins daniel_hozac: it wasn't in inet_csk_bind_conflict and ipv6_rcv_saddr_equal 1190910423 M * daniel_hozac ipv6_rcv_saddr_equal has the same code, only different. 1190910454 M * baggins and that difference is killing my server :/ 1190910491 M * daniel_hozac do you have a reproducer? because it works fine here... 1190910511 M * fb daniel_hozac: is there a way to mount --bind a tree with another owner and/or user rights? 1190910525 M * daniel_hozac and seriously, ipv6_only_sock doesn't do anything crash-worthy. 1190910538 M * daniel_hozac fb: no. 1190910551 M * baggins sorry, it just happens, see http://paste.linux-vserver.org/6779, the problem is call to ipv6_only_sock via inet_csk_bind_conflict 1190910568 M * fb hm, so i must remember this feature srom solaris, and that's why i cannot google it :) 1190910573 M * fb thanks 1190910605 M * daniel_hozac i haven't seen any addr2line for it. 1190910678 M * baggins I seen the disasm code, and it's blatantly clear the oops is caused by ipv6_only_sock 1190910696 M * daniel_hozac which part? 1190910711 M * baggins the code tries to check a bitfield which isn't there 1190910737 M * daniel_hozac that's only if the address family is PF_INET6. 1190910753 M * baggins f6 40 52 20 testb $0x20,0x52(%rax) 1190910796 Q * ema Quit: leaving 1190910868 M * daniel_hozac and what's %rax? 1190910895 J * bonbons ~bonbons@2001:960:7ab:0:20b:5dff:fec7:6b33 1190910896 M * baggins struct sock *sk2 1190910983 J * _jmcaricand_zzz ~jmcarican@d77-216-208-29.cust.tele2.fr 1190910998 M * daniel_hozac so wait, is that the instruction you claim is oopsing? 1190911050 M * daniel_hozac because if so... why is sk2 NULL? 1190911053 M * baggins kernel claims it's the instruction that is oopsing 1190911106 M * daniel_hozac and why didn't it oops earlier? 1190911224 M * daniel_hozac it should've oopsed much earlier, if %rax indeed is sk2. 1190911460 Q * _jmcaricand_zzz Quit: KVIrc 3.2.4 Anomalies http://www.kvirc.net/ 1190911469 J * opuk ~kupo@c213-100-138-228.swipnet.se 1190911611 M * baggins I did some digging, that ipv6 check was somewhere there, indirect tho 1190911681 Q * pmenier Quit: pmenier 1190912485 M * julien-vserver daniel_hozac: daniel_hozac: echo 128 > /proc/sys/vserver/debug_misc and redo it. <- was it for me ? are we agree that I attach strace to X ? 1190912520 M * daniel_hozac you don't really need strace with that. 1190912563 M * julien-vserver I will see what I "want" in startx output ? 1190912575 M * julien-vserver ok, I try 1190912579 M * daniel_hozac dmesg should have the data, if it's indeed vserver blocking it. 1190912615 M * julien-vserver ok 1190912634 J * Piet ~piet@tor.noreply.org 1190912731 M * hparker daniel_hozac: any idea why openvcp would error out like this? http://rafb.net/p/wisA7I74.html 1190912742 M * daniel_hozac they don't define tag_t. 1190912768 M * hparker ty 1190913551 M * julien-vserver I'm tired, trying to debug something from debian unstable by ssh, how crazy am I... machine freezes... 1190914143 M * julien-vserver ok, I stop for today, I will restart to search something but I need to prepar my ibook for this new semester, I writed all I did so I think I will come back quickly to the same state 1190914321 Q * JonB Ping timeout: 480 seconds 1190914887 Q * Piet Ping timeout: 480 seconds 1190915173 Q * meandtheshell Quit: Leaving. 1190915499 J * Piet ~piet@tor.noreply.org 1190915583 M * jescheng hello 1190916013 Q * jescheng Remote host closed the connection 1190916027 J * jescheng ~jescheng@proxy-sjc-1.cisco.com 1190916038 M * jescheng by default, can we insert kernel modules in vserver? 1190916056 M * tokkee jescheng: No. 1190916062 M * daniel_hozac there's only one kerenl. 1190916100 M * julien-vserver I'm not sure that's what he asked 1190916147 M * jescheng can we set this capabilities flag "SYS_MODULE" to allow it 1190916158 M * daniel_hozac yes, but that's not something you want to do. 1190916175 M * daniel_hozac mostly because you likely won't have the correct kernel modules in the guest. 1190916217 M * jescheng assuming we built the modules for the same kernel... 1190916269 M * daniel_hozac you realize that giving the guest CAP_SYS_MODULE will void any and all security features provided by Linux-VServer, right? 1190916348 M * jescheng yes, i understnad, just wanted to check if it's possible to do 1190916680 M * daniel_hozac everything is possible if you don't care about security. 1190916799 Q * coderanger_ Quit: coderanger_ 1190916891 M * jescheng i see.. can this flag be set per guest (i.e. if we can identify certain guest to be trusted) 1190916913 M * daniel_hozac capabilities? of course. 1190916969 M * jescheng ok great. once it's added to the kernel..it's going to be available to all guests rite...since it's the same kernel 1190916979 M * daniel_hozac yes. 1190917006 M * jescheng ok. ty 1190917078 J * JonB ~NoSuchUse@kg0-231.kollegiegaarden.dk 1190917437 J * _jmcaricand_zzz ~jmcarican@d77-216-208-29.cust.tele2.fr 1190917558 Q * julius__ Quit: Verlassend 1190917562 J * julius__ ~julius@p57B266E4.dip.t-dialin.net 1190917897 Q * JonB Ping timeout: 481 seconds 1190917956 J * JonB ~NoSuchUse@kg1-20.kollegiegaarden.dk 1190918859 Q * JonB Quit: This computer has gone to sleep 1190919091 N * Bertl_zZ Bertl_oO 1190919171 J * Piet_ ~piet@tor.noreply.org 1190919210 J * ema ~ema@rtfm.galliera.it 1190919552 Q * Piet Ping timeout: 480 seconds 1190919693 J * Supaplex ~e@166-70-62-194.ip.xmission.com 1190919729 M * Supaplex humm looks like my fc7 (guest) install is incomplete. I have no: which, rpm, or yum. 1190919734 A * Supaplex checks the wiki 1190919793 M * Supaplex http://linux-vserver.org/Installing_a_Fedora_7_vserver_guest 1190919796 M * Supaplex nice... 1190919800 M * Supaplex it's empty :P 1190919831 M * fb reload to update cache 1190919859 M * fb hm 1190919867 M * fb it's empty %-) 1190919879 A * Supaplex hears echos of himself 1190919885 M * Supaplex :) hehe 1190919909 M * Supaplex I have a new pc to go pickup, soooo... I'll bbl. 1190922079 Q * _jmcaricand_zzz Quit: KVIrc 3.2.4 Anomalies http://www.kvirc.net/ 1190923743 Q * jescheng Quit: Leaving 1190924618 J * coderanger_ ~coderange@ae-lally-green-77.dynamic2.rpi.edu 1190925035 J * Moo ~troy@shells195.pinchaser.com 1190925049 A * Moo moos while entering. 1190925053 N * Moo MooingLemur 1190925125 Q * julius__ Remote host closed the connection 1190925125 Q * ktwilight Read error: Connection reset by peer 1190925151 M * MooingLemur Does anyone here use nfs mounts for vserver roots, where they're only mounted in the vserver context? 1190925273 J * ktwilight ~ktwilight@197.220-66-87.adsl-static.isp.belgacom.be 1190925648 Q * bonbons Quit: Leaving 1190925753 M * Supaplex MooingLemur: I should be soon for a /net/nfs-dump-box/{backups,sbin} set of exports 1190925768 M * Supaplex soon/few days/1w or so 1190926104 Q * Punkie Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.7/2007091417] 1190926345 J * Aiken ~james@ppp121-45-249-108.lns2.bne4.internode.on.net 1190927783 Q * ruskie Write error: connection closed 1190927962 J * ruskie ruskie@ruskie.user.oftc.net 1190928873 J * arekm arekm@carme.pld-linux.org 1190928876 M * arekm hello 1190928886 M * daniel_hozac hi 1190928985 M * arekm http://pastebin.com/m5f7f6115 2.6.22.9 + some vserver 2.3 snap, known maybe? 1190929003 M * daniel_hozac no. 1190929102 M * daniel_hozac can you reproduce it? 1190929163 M * arekm yes, it requires booting that kernel, waiting few minutes and oops happens 1190929258 M * arekm it's patch-2.6.22.6-vs2.3.0.24.diff 1190929264 M * daniel_hozac what does that guest do? 1190929312 M * daniel_hozac (the one that stops itself) 1190929357 M * arekm no idea since I can't do anything when this happens and machine reboots 1190929403 M * arekm anyway all 7 guests here doing typical things, httpd, sshd, ftpd etc 1190929421 M * daniel_hozac so why is one of them stopping itself? 1190929492 M * arekm why do you say it's stopping itself? (aka which part of log indicates that) 1190929537 M * daniel_hozac well, i assume you're not stopping it, right? 1190929545 M * arekm yes 1190929549 M * daniel_hozac "RIP [] free_vx_info+0x78/0xa0" 1190929592 M * sid3windr requiescat in pace 1190929608 Q * coderanger_ Quit: coderanger_ 1190929838 J * yarihm ~yarihm@84-75-130-73.dclient.hispeed.ch 1190930374 M * arekm the weird thing is that similar oops happened for me with older kernels, too 1190930376 M * arekm http://lists.pld-linux.org/mailman/pipermail/pld-kernel/2007-July/001409.html 1190930392 M * arekm and http://lists.pld-linux.org/mailman/pipermail/pld-kernel/2007-July/001403.html 1190930427 M * daniel_hozac and that's with a completely vanilla kernel? 1190930431 M * arekm 2.6.17 + some vserver 2.1 is on the other hand very stable 1190930444 J * coderanger_ ~coderange@pepe-30.dynamic2.rpi.edu 1190930458 M * arekm not vanilla unfortunately 1190930468 M * daniel_hozac try vanilla then. 1190930852 Q * ema Quit: leaving 1190930995 M * julien-vserver as I need to reinstall my system for my studies I will give a try for debian unstable after testing echoing 128 to debug_misc 1190931036 M * julien-vserver read you later, and thanks for your time 1190931044 Q * julien-vserver Quit: Leaving. 1190931150 M * dowdle I'm trying to create new VPSes... with the instructions on the Linux-VServer CentOS 5 Install page... and I'm getting the following error... any ideas? 1190931151 M * dowdle rpm-fake-resolver: vc_ctx_migrate(): No such process 1190931151 M * dowdle rpm-fake.so: failed to initialize communication with resolver 1190931151 M * dowdle rm: cannot remove directory `/etc/vservers/.defaults/vdirbase/mail2/proc': Device or resource busy 1190931151 M * dowdle rm: cannot remove directory `/etc/vservers/.defaults/vdirbase/mail2/.rpmdb': Device or resource busy 1190931182 M * daniel_hozac how did you build the utils? 1190931289 M * dowdle daniel_hozac: I'm using the utils from your Linux-VServer CentOS 5 repo... totally binary install. 1190931301 M * dowdle I was able to create a guest yesterday just fine. 1190931334 M * dowdle And just seconds ago I accidentally deleted the filesystem of my running VPS... when trying to delete the aborted/problem dirs... so now I have a VPS that is kinda running but not. 1190931338 M * dowdle That I can't stop. 1190931339 M * dowdle Opps. 1190931357 M * daniel_hozac vkill is your best buddy. 1190931419 Q * dna Quit: Verlassend 1190931449 M * daniel_hozac what context did you use? 1190931461 M * dowdle Not without a man page it isn't. 1190931477 M * dowdle 4040 as in the example... and then I just increased it by one for each additional VPS I was going to create. 1190931488 M * daniel_hozac as with everything else in util-vserver, --help. 1190931588 M * dowdle daniel_hozac: What am I supposed to be vkilling? 1190931599 M * daniel_hozac the guest? 1190931654 M * dowdle And how do I refer to the guest? 1190931666 M * daniel_hozac by the context id? 1190931708 M * dowdle [root@comp2 dbase]# vkill 4040 1190931708 M * dowdle vkill: vc_ctx_kill(): No such process 1190931726 M * daniel_hozac 4040 is an xid, not a pid, no? 1190931792 M * dowdle Well, I don't have any VPSes running... but I now have a bunch of directory structures that are "busy". My VPS that was running... that I accidentally nuked the filesystem of... isn't running anymore. 1190931802 M * dowdle xid = context id? 1190931855 M * daniel_hozac yes. 1190931871 M * daniel_hozac so you've already killed all the processes? 1190931890 M * dowdle I haven't killed any processes. 1190931910 M * dowdle Perhaps the VPS that was running died on its own... after loosing most of the files that make it up. 1190931915 M * daniel_hozac then what makes you say it's not running anymore? 1190931943 M * dowdle In any event... I don't have any VPSes running (vserver-stat shows none) and I can't create any... and the broken creation attempts created dirs I can't get rid of now. 1190931980 M * dowdle Maybe my attempts to use vkill did work... but the error message threw me off. 1190931997 M * daniel_hozac so vps faux doesn't show any processes not belonging to context 0 or 1? 1190932056 M * dowdle (wow, I learned a new command)... no... everything is owned by 0 and 1 1190932084 M * daniel_hozac that's odd. 1190932113 M * daniel_hozac so rm -fr /vservers/ still returns -EBUSY? 1190932159 M * dowdle Yes. 1190932192 M * daniel_hozac and you don't have any lingering processes from the build commands that shouldn't be there? 1190932194 M * dowdle BTW... vps --help on an ancient (2.4.x kernel) returns a bunch of help info. vps --help on my new install... only returns one line... same goes for --help on all of the commands I've tried. Is that normal? 1190932239 M * dowdle No sir... not that I can identify 1190932252 M * daniel_hozac vps is just a ps wrapper, it doesn't take options. 1190932313 M * daniel_hozac the only time --help is a one-liner is if the command is sufficiently self-explainatory not to warrant any description of the options 1190932321 M * dowdle ok. 1190932332 M * dowdle That was fancy how you called me stupid. :) 1190932338 M * dowdle hehe 1190932342 M * daniel_hozac :) 1190932373 M * dowdle I guess I'll reboot the hn and see if Jesus loves me. 1190932497 N * Bertl_oO Bertl_zZ 1190932582 Q * _gh_ Quit: Client exiting 1190932617 Q * igraltist Ping timeout: 480 seconds 1190932655 M * dowdle So, where do I put the linux32 in my VPS creation statement? Anywhere after the build keyword? 1190932674 M * daniel_hozac you prepend the command with it. 1190932683 M * daniel_hozac i.e. linux32 vserver ... 1190932704 M * dowdle DUH. Ok. 1190932708 M * dowdle That prepend. 1190932975 M * dowdle What provides /vservers/.pkg? 1190933079 M * dowdle Nevermind. a mkdir made things happy. 1190933251 Q * arekm Quit: leaving 1190933271 J * arekmx ~arekm@chello089076024040.chello.pl 1190933304 A * arekmx reboots to 2.6.22 + .9 patch + vserver 2.3.0.24 1190933355 M * dowdle Ok, rebooting made everything happy again and I'm back to just enough info to hurt myself. :) 1190933868 M * dowdle What value goes in the personality again? (I'm taking notes this time) i386? linux32? 1190933907 M * dowdle Ah the FAQ says linux_32bit 1190933936 M * dowdle Hmmm, that doesn't work. 1190933990 M * dowdle Ah i686 1190934036 M * dowdle Nope, that doesn't work either. 1190934063 M * dowdle Ok, linux32 it is. 1190934240 P * dowdle Gotta catch the bus nome now. Thanks for your patience and help. 1190934348 Q * besonen_mobile_ Read error: Connection reset by peer 1190934348 J * besonen_mobile_ ~besonen_m@71-220-225-14.eugn.qwest.net 1190934457 M * arekmx daniel_hozac: here it is http://pastebin.com/m9df408a 1190934756 Q * yarihm Quit: Leaving 1190934899 J * arekm arekm@carme.pld-linux.org 1190934909 Q * arekmx Quit: now as arekm 1190935041 A * arekm goes to sleep now but leaves session for saving comments about the issue 1190936051 Q * ruskie Remote host closed the connection 1190936133 Q * AndrewLee Quit: leaving 1190936238 J * ruskie ruskie@ruskie.user.oftc.net 1190936370 J * AndrewLee ~andrew@flat.iis.sinica.edu.tw 1190937033 Q * coderanger_ Quit: coderanger_