1150588805 M * Bertl and again, it has nothing to do with the _limits_ 1150588814 A * Hollow head moves in mysterious mays 1150588833 M * Hollow fingers too 1150588838 M * Bertl :) 1150588839 M * Nam any more help for me Bertl? 1150588856 M * Hollow it's already quite late too 1150588884 M * Bertl the /sbin/vcontext and /lib/util-vserver/save_ctxinfo do exist on the host too, I presume? 1150588891 M * Nam yup 1150588894 M * Hollow Bertl: guess i looked for the special magic, but it's plain and simple :p 1150588910 M * Bertl Hollow: yeah, well then off to bed for you? 1150588912 M * Hollow and maybe i still confused it a bit with min limits 1150588926 M * Hollow not yet 1150588929 M * Bertl Nam: what about /sbin/init in the guest? 1150588969 M * Hollow i need to change my uptimes, so best i start now ;) 1150588982 M * Nam /etc/vservers/hosting/vdir/sbin/init yup 1150588985 M * Bertl Nam: ah, and could you upload the output of testme.sh and 'vserver-info - SYSINFO' please? 1150588994 M * Nam testme.sh? 1150589035 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/ 1150589119 M * Nam http://pastebin.ca/66689 1150589129 M * Nam testme.sh was all successful 1150589240 M * Bertl looks good 1150589263 M * Bertl could you try to build a guest via util-vserver (from the network)? 1150589330 M * Bertl something like this: 1150589331 M * Bertl vserver foo build -m debootstrap --hostname vs.foo.org --netdev eth0 --interface 192.168.3.1/21 --context 42 -- -d sarge -m ftp://ftp.debian.org/debian 1150589349 M * Nam i am using slackware 1150589355 M * Bertl that should be fine 1150589382 M * Nam i want a slackware guest 1150589390 M * Nam i personally can't stand debian 1150589421 M * Nam thanks for the great support so far btw, i really appreciate your help 1150589422 M * Bertl well, would be interesting to see if that works, no? 1150589433 M * Bertl you're welcome! 1150589511 M * Nam while that command is running... are you one of the developers for this project? 1150589536 M * Bertl yes, I am working on the kernel ... 1150589555 M * Nam cool, good to see your know and support the project you work on well ;) 1150589594 M * Nam on a side note... is vserver compatible with the slackware package management system, or only deb/rpm? 1150589601 M * Bertl I'm using it for my purposes so I have to know a few things ... 1150589618 M * Bertl I haven't seen anybody add support for slackware yet 1150589643 M * Bertl but basically it should be quite simple if you know slackware's management system 1150589682 M * Nam i am building a entire management system which is going to use this vserver project as it's virtualization 1150589728 M * Bertl cool 1150589733 M * Nam it's under development right now, but we hope to be going public soon, i'll talk to you and the rest of the team when that times comes near, and see about establishing some kind of relationship 1150589762 M * Bertl will it be open source? 1150589770 M * Nam i do know slackwares package management system well... i just don't know how this project managed the project management yet... to many things my mind is on right now 1150589787 M * Nam for starters.... probably not... but that is the goal yes 1150589807 M * Bertl i.c. 1150589813 M * Nam it's only going to be available through our company until we feel comfortable with the code level to opensource it 1150589837 M * Nam i'm security anal, so, i'll be having me and a team go through all the final code with a fine tooth comb 1150589847 M * Nam command is still running btw 1150589864 M * Piet could someone refer me to any good comparison of linux-vserver vs xen? 1150589871 M * Bertl Nam: does it make/show some progress? 1150589889 M * Nam it's on the P's right now 1150589891 M * Bertl Piet: we have a paper from princeton linked on the wiki 1150589924 M * Bertl but to make it short, Xen does strict resource assignments and has a kernel for each guest running 1150589948 M * Bertl Linux-VServer virtualizes on the kernel/app level and is all about resource sharing and performance 1150589961 M * Nam http://www.cs.princeton.edu/~mef/research/vserver/paper.pdf 1150589971 M * Bertl but you can combine both projects quite nicely 1150589972 M * Piet thanks 1150589989 M * Bertl i.e. have a Linux-VServer kernel running in DomU 1150590014 M * Bertl (and for example a bsd kernel in another one) 1150590016 M * Nam yea... i was thinking of using Xen... but then i thought to myself... i don't want to run anything other then linux... so, no point really for the my needs 1150590043 M * Piet ok, thanks you 1150590053 M * Bertl you're welcome! 1150590119 M * Nam command finished 1150590125 M * Nam save_ctxinfo: execv(): No such file or directory 1150590127 M * Nam same thing 1150590204 M * Nam anyway we can debug this more to find out what file or directory save_ctxinfo is trying to access? 1150590212 Q * matti Ping timeout: 480 seconds 1150590279 M * Bertl yep, let me search the archives, I have a feeling we already had that 1150590290 M * Nam if i put in the follow command... what should the responce be? 1150590293 M * Nam /sbin/chbind --silent --nid 42 --ip 192.168.1.77/24 -- /lib/util-vserver/exec-ulimit /etc/vservers/foo/ulimits /sbin/vcontext --create --silent --xid 42 -- /sbin/vnamespace --set -- /sbin/vlimit --dir /etc/vservers/foo/rlimits --missingok -- /sbin/vsched --xid self -- /sbin/vuname --xid self --dir /etc/vservers/foo/uts --missingok -- /sbin/vuname --xid self --set -t context=/etc/vservers/foo -- /sbin/vattribute --set --secure --flag default --flag fakeinit -- /lib 1150590293 M * Nam /util-vserver/save_ctxinfo /etc/vservers/foo /usr/bin/env -i -- /sbin/vcontext --migrate-self --endsetup --chroot --silent -- /etc/init.d/rc 3 1150590305 M * Nam it's all one line 1150590315 M * Bertl well, it basically starts the guest 1150590325 M * Nam i am getting this as a responce 1150590326 M * Nam save_ctxinfo: open("/var/run/vservers/foo"): File exists 1150590342 M * Bertl yep, it thinks it is already started 1150590354 M * Bertl try to stop it and/or remove the file 1150590373 M * Nam root@Mary:/var/run/vservers# vserver foo status 1150590373 M * Nam Vserver 'foo' is stopped 1150590395 M * Bertl let's check the following things: 1150590407 M * Nam hmm... 1150590414 M * Nam if i do vserver foo stop 1150590423 M * Nam then run that command i pasted above... i get this... 1150590429 M * Nam vcontext: execvp("/etc/init.d/rc"): No such file or directory 1150590457 M * Bertl which is similar to the previous one (with /sbin/init) 1150590464 M * Bertl try to stop it again and do: 1150590477 M * Bertl ..... -- /bin/bash 1150590498 M * Bertl (I assume you have /bin/bash in the guest) 1150590541 M * Nam same thing 1150590547 M * Nam vcontext: execvp("/bin/bash"): No such file or directory 1150590551 M * Nam but it's there 1150590573 M * Bertl okay, so let's heck the permissions of this file and the directories above 1150590635 M * Bertl i.e. ls -lagd /vservers /vservers/foo /vservers/foo/bin 1150590653 M * Bertl and the same with lsattr -d 1150590656 M * Nam well... the vserver build command worked, it should have the proper permissions from the build shouldn't it? 1150590659 M * Bertl ah, and let's add / too 1150590673 M * Nam drwxr-xr-x 12 root 272 2006-06-17 12:08 /vservers/ 1150590673 M * Nam drwxr-xr-x 20 root 480 2006-06-17 12:17 /vservers/foo/ 1150590673 M * Nam drwxr-xr-x 2 root 1992 2006-06-17 12:17 /vservers/foo/bin/ 1150590683 M * Bertl basically yes, but we do not know about / and /vservers 1150590713 M * Nam root@Mary:/vservers# lsattr -d /vservers /vservers/foo /vservers/foo/bin 1150590713 M * Nam lsattr: Inappropriate ioctl for device While reading flags on /vservers 1150590716 M * Nam same for all of them 1150590736 M * Bertl hmm, what filesystems are those? 1150590819 M * Nam / is ext3, and /vservers is reiser 1150590848 M * Bertl lsattr shows the ext2/3 attributes on filesystems 1150590861 M * Bertl it's somewhat unlikely that ext2/3 doesn't allow that :) 1150590878 M * Bertl for reiserfs, they have to be enabled using attribs 1150590882 M * Bertl (as mount option) 1150590968 M * Bertl hm, attrs actually 1150591009 M * Nam both / and /vservers are set to defaults 1150591026 M * Bertl means you get only dashes? 1150591060 M * Nam /dev/hdd1 /vservers auto defaults 1 2 1150591061 M * Bertl ah, defaults as the fstab mount option 1150591067 M * Nam /dev/hda3 / ext3 defaults 1 1 1150591093 M * Nam ------------- /root 1150591093 M * Nam ------------- /opt 1150591093 M * Nam lsattr: Inappropriate ioctl for device While reading flags on /vservers 1150591110 M * Bertl ah, okay, so / should work too 1150591116 M * Nam if i do lsattr in / some things work, some don't 1150591140 M * Bertl well, as I said, you want to mount the reiserfs with the attrs option 1150591152 M * Nam /dev, /sys, /proc, and /vservers get the "lsattr: Inappropriate ioctl for device While reading flags on" error 1150591169 M * Nam how do i do that? 1150591174 M * Bertl which are virtual filesystems so that is fine 1150591182 M * Nam add ,attrs? 1150591187 M * Bertl simply by unmounting /vservers and adding the attrs to the fstab 1150591195 M * Bertl then mounting it again 1150591378 M * Nam i have to reboot it 1150591385 M * Nam just take a moment 1150591506 M * Nam there we go 1150591517 M * Bertl looks like we had at least 3 incidents which show the same 'strange' behaviour, for two systems I can confirm that they are slackware 1150591531 M * Nam ------------- /vservers 1150591531 M * Nam ------------- /vservers/foo 1150591531 M * Nam ------------- /vservers/foo/bin 1150591543 M * Bertl the interesting thing is, it could not be resolved :/ 1150591571 M * Nam hehe... can't be the distribution... linux is the same, it's just in a different configuration, anything can be fixed ;) 1150591591 M * Bertl yeah, I suspect the env part to be eviφ 1150591593 M * Nam i've been using slackware for about 10 years now, i have always gotten everything to work on it 1150591595 M * Bertl *evil even 1150591626 M * Bertl let's try the 'long line' without the env part 1150591684 M * Bertl i.e. remove the '/usr/bin/env -i --' 1150591735 M * Nam remove everything after env? or just remove the '/usr/bin/env -i --' from the string? 1150591747 M * Bertl just the small part 1150591768 M * Nam same error 1150591769 M * Nam vcontext: execvp("/etc/init.d/rc"): No such file or direcvcontext: execvp("/etc/init.d/rc"): No such file or directory 1150591769 M * Nam tory 1150591772 M * Nam woops 1150591780 M * Nam didn't mean to past multiple lines 1150591788 M * Bertl np 1150591814 M * Nam it's like it's not going into the vserver to access the file /etc/init.d/rc 1150591840 M * Bertl well, if that was true, the /bin/bash version would have started the host's bash 1150591907 M * Nam not if for some reason... it's using the wrong folder 1150591931 M * Bertl well, let's check the symlinks in the config for a start 1150591982 M * Nam ls vdir/ 1150591982 M * Nam bin/ boot/ dev/ etc/ home/ initrd/ lib/ media/ mnt/ opt/ proc/ root/ sbin/ srv/ sys/ tmp/ usr/ var/ 1150592006 M * Nam oot@Mary:/etc/vservers/foo# ls vdir/etc/init.d/rc 1150592006 M * Nam vdir/etc/init.d/rc* 1150592015 M * Nam it's there 1150592049 M * Bertl what about the run symlink? 1150592090 M * Nam run -> /var/run/vservers/foo 1150592135 M * Nam hmm... is the value of run supposed to be the context id? 1150592163 M * Bertl yep 1150592168 M * Nam ok 1150592171 M * Nam then it's good 1150592183 M * Bertl you did upload the output of testme.sh somewhere, yes? 1150592214 M * Nam i sent you the link to that a while ago... it was below the result of the vserver-info or what ever it was 1150592226 M * Nam it's was all successful 1150592239 M * Bertl do you have the url at hand? 1150592284 M * Bertl somehow closed that window ... 1150592295 M * Nam http://pastebin.ca/66689 1150592323 M * Bertl ah, maybe the grsec part is to blame ... 1150592335 M * Nam that's a possibility 1150592351 M * Bertl could we try with a vserver only kernel (just to make sure)? 1150592351 M * Nam you know what options would cause it? 1150592385 M * Bertl no idea, but I'm pretty sure there are plenty of possibilities 1150592387 M * Nam i'd prefer to keep grsecurity.... i'll go and just disable all the grsecurity options in the kernel 1150592393 M * Nam see if that changes it 1150592406 M * Nam if that doesn't... then i'll try just a vserver patch 1150592415 M * Bertl okay, that's fine with me 1150592436 M * Nam hehe... hope this doesn't get annoying, but thanks again for all your help 1150592463 M * Bertl no, not at all, I see it as debugging ... you're quite welcome! 1150592482 M * Nam yup... i understand how that goes... i'm the same way 1150592503 M * Nam once we go public and start making lots of money, i'll make sure to talk my boss into doing a nice sized donation to the project 1150592509 M * Nam we are building a data center 1150592564 M * Nam i'm building a web based interface to manage everything... not just vservers on a local machine, but managing an entire office, or multiple offices and all the systems 1150592723 M * Nam how old are you? 1150592736 M * Bertl almost 36 1150592748 M * Nam coo.. i'm 24 1150592760 M * Nam got any kids? 1150592767 M * Bertl not yet 1150592784 M * Nam i'm on my second right now.. probably all we are going to have 1150592849 M * Nam hope it's a boy.. we already have a girl... it sucks in our area.. because of some religious beliefs by extremists out here who kill female babies.. the local medical doctors decided not to tell anyone the sex of their babies 1150592904 M * Bertl you are located where? 1150592911 M * Nam Canada BC 1150592973 M * Nam u? 1150592981 M * Piet wow, this still happens... 1150592988 M * Bertl Nam: Austria, Europe 1150592996 M * Nam cool 1150593003 M * Nam yes, i know, its sad 1150593080 M * Piet BC stands for british columbia? 1150593107 M * Nam Piet: it's worse in other areas of the world... because in canada and the us we have been raised to believe you shouldn 1150593131 M * Nam 't kill a baby like that, but in orther parts of the world, if you have a girl,,, you just becamse poor 1150593134 M * Nam yup 1150593147 M * Nam i love all the mountains and woods here in BC 1150593155 M * Nam i'm a big camper 1150593158 M * Piet i had not expected something like this to happen in canada really 1150593189 M * Nam yup, the medical industry is corrupt everywhere, even canada 1150593191 M * Piet guess i was wrong ;) 1150593199 M * Piet right 1150593253 M * Piet do you know a number? likw how many babies are killed that way a year? 1150593294 M * Nam did you ever hear about Bayer purposly selling aids contaminated drugs overseas causing hundreds of thousands of deaths, we still are actually feeling the effects of it, because it being passed into kids, and stuff, the effects of that are still spreading 1150593304 M * Nam nope, no idea 1150593331 M * Nam the thing is... people get a test... if it's a girl, they make themselves have a miscarage 1150593343 M * Nam so, it's hard to proove or keep track of 1150593426 M * Piet obviously. that should cause some trouble with reproduction at some point ;) 1150593465 M * Nam remember, in many traditional famlies, a family of girls have to pay a dowlry still, they have to offer everything they can to the family of the males in order to have the marage accepted 1150593491 M * Nam unless your crazy rich, if you have a girl, your family will be poor shortly after 1150593530 M * Nam damn it!! 1150593534 M * Piet i know this, but i was thinking this to be more of an oriental habit. 1150593543 M * Nam i just recompiled the kernel, with 0 grsecurity support, and still same error 1150593559 M * Piet s/habit/culture/ 1150593571 M * Bertl Nam: so next step vanilla + vserver patch :) 1150593572 M * Piet anyway, we're going widely offtopic ;) 1150593572 M * Nam not only oriental, but east indian as well... we have lots of both all over the areas here 1150593621 M * Piet i see, i hope this is gonna change soon 1150593642 M * Nam Bertl: yup, i'm downloading and i'll compile it 1150593676 M * Bertl personally I would opt for an userspace issue, but better to make sure here 1150593721 M * Nam Piet: it's going to get much worse before it gets any better 1150593723 M * Nam ;) 1150593819 M * Piet press the fast forward button then ;) 1150593846 M * Nam hehe, i wish 1150593861 M * Nam i wish i had that remote from the new movie "Click" hehe 1150593951 Q * mire Remote host closed the connection 1150594032 M * Piet :) 1150594059 Q * doener Ping timeout: 480 seconds 1150594207 J * mire ~mire@206-166-222-85.COOL.ADSL.VLine.Verat.NET 1150594445 M * Nam compiling now 1150595264 M * Nam rebooting with new kernel now 1150595270 M * Bertl okay 1150595376 M * Nam guess what 1150595382 M * Nam save_ctxinfo: execv(): No such file or directory 1150595386 M * Nam same fricken thing 1150595399 M * Bertl well, that is good news too 1150595417 J * jkl eric@71.56.216.223 1150595430 M * Bertl after all it means that the grsec kernel should be fine 1150595430 M * Nam oooo.... 1150595434 M * Nam yup 1150595440 M * Nam i just tried something different 1150595457 M * Nam i tried starting one of the old vservers with the old config 1150595463 M * Bertl and? 1150595473 M * Nam error was this... "chcontext: tools were built without legacy API support; can not continue" 1150595482 M * Nam ah 1150595484 M * Nam actually 1150595486 M * Nam hehe, makes sence 1150595491 M * Bertl yeah 1150595494 M * Nam hehe, duh 1150595523 M * Nam well... can you think of where to go from here? 1150595524 M * Bertl well, let's check a few things with userspace 1150595550 M * Bertl for one thing, I'd like to see the tools compiled with a different gcc version 1150595559 M * Bertl actually the tools _and_ dietlibc 1150595575 M * Bertl can this be arranged? 1150595587 M * Bertl e.g. 3.3.5 or something like that? 1150595625 M * Nam i'm using glibc-2.3.6 ... something wrong with that? 1150595636 M * Nam it's the latest one in slackware-current 1150595641 M * Bertl gcc not glibc 1150595678 M * Nam ah... 1150595697 M * Nam 3.4.6 i'm on 1150595702 M * Nam something wrong with it? 1150595900 M * Bertl well, I know it works with 3.3.5/6 and 4.0.xsomething 1150595944 M * Bertl so it might be worth a try with one of them, especially as a fellow user reported issues which 'might' be related to the gcc version (which was also 3.4.something) 1150595945 M * Nam i'm grabbing a copy of the 10.2 gcc, which is 3.3.6 1150595949 M * Nam i'll downgrade 1150595958 M * Bertl yeah, that version sounds good 1150595969 M * Bertl as I said, make sure to rebuild the dietlibc too 1150595971 M * Nam if those people where using -current 1150595984 M * Bertl you are working on x86 arch, yes? 1150595989 M * Nam what about glibc?... will i have to do that? 1150595997 M * Nam x86 yes 1150596013 M * Bertl glibc is not much involved 1150596025 M * Nam i just upgraded everything last night, hehe 1150596025 M * Bertl the tools use the dietlibc to make them secure 1150596108 M * Nam while that's dling... you got any links or advice about getting the vserver commands using the slackware package management instead of rpm/deb? 1150596195 M * Piet good nite 1150596202 M * Nam night Piet 1150596211 M * Piet :) 1150596217 Q * Piet Quit: :tiuQ 1150596217 M * Bertl good nigt Piet! 1150596237 M * Bertl Nam: well, the basic 'elements' for building a guest are: 1150596256 M * Bertl - the ability to 'bootstrap' the distro into a directory 1150596284 M * Bertl - some interface for 'manipulating' the package database from outside 1150596299 J * doener ~doener@p5487782F.dip.t-dialin.net 1150596358 M * Bertl as example, the debootstrap method downloads debootstrap and uses that, together with the url and arch stuff to build the guest in a dir 1150596395 M * Bertl different example is the vrpm wrapper, which allows to do rpm 'inside' a guest 1150596465 M * Nam ok... so, the bootstrap is a single file which contains information for installing the base system? 1150596562 M * Bertl well, this was the debian example, and debootstrap in debian is a single executable which can install a debian system into a dir 1150596586 M * Nam ah... like a single bash script with all the commands? 1150596595 M * Bertl for rpm this is usually done via apt-rpm or yum 1150596624 M * Nam where can I find an example bootstrap file do you know? 1150596674 M * Bertl as I said, for debian this is a tool (which knows all the details) 1150596694 M * Bertl I don't know how slackware is installed into a dir 1150596716 M * Bertl how is it usually installed? is there a network install method? 1150596716 M * Nam yup, i understand that part... so far, i have had to make all my own tools manually 1150596745 M * Nam it would be nice to integrate it into the vserver utils 1150596767 M * Nam typically installed via filesystem/cd/dvd/nfs 1150596792 M * Bertl okay, let's assume we have a CD, what happens to install it? 1150596802 M * Nam on the main system that the vserver kernel is running on, i'll be keeping an archive of every distro we are going to be using 1150596814 M * Nam so, i can custom code it to access those files 1150596837 M * Bertl check /usr/lib/util-vserver/vserver-build.debootstrap as example 1150596846 M * Nam to install it, you boot to a prompt, you log in, partition the disks, then run "setup" which goes through and installs the packages 1150596866 M * Bertl okay, so you want to start at the 'setup' phase 1150596877 M * Bertl i.e. after the fielsystems are created 1150596894 M * Bertl can you advise the setup tool/script to put it into a directory? 1150596916 M * Nam for that, you would use a series of installpkg commands 1150596946 M * Bertl okay, that would be possible too, we use that for the 'plain' rpm install method 1150596957 M * Nam if you run "installpkg slackwarebasedir/slackware/a/*.tgz" that would install a basic set of the system 1150596957 M * Bertl what about dependancy resolving? any tool for that? 1150596987 M * Nam nope, the package management system doesn't handle dependancies 1150597023 M * Bertl hmm, so who does? 1150597042 M * Nam hehe, user 1150597050 M * Nam typically, slackware users know what they are doing 1150597056 M * Bertl ah, okay, so that is fine then 1150597075 M * Bertl in this case it should be simply a 'variant' of the rpm install method 1150597088 M * Nam so, dependancies have never been a problem, plus, slackware usually puts the dependancy libs into packages 1150597091 M * Bertl just with a bunch of tgz's instead of rpms 1150597197 M * Nam ahh..... 1150597234 M * Nam ok. so... vserver-build is just a bash script, which is calling the appropriate vserver-build.functions.* and vserver-build.* files 1150597262 M * Bertl yep, it does this depending on the -m 1150597299 M * Nam what about when using "vserver pkg install "? 1150597503 M * Bertl this is a little trickier 1150597521 M * Bertl it basically utilizes the v* wrapper 1150597531 M * Bertl i.e. if the tool is rpm, we have vrpm 1150597558 M * Bertl in your case that would probably be vinstallpkg 1150597599 M * Bertl the vserver pkg ... just calls that based on the guest setup 1150597726 M * Nam vrpm and vdeb part of this util-vserver, or part of their own package systems? 1150597744 M * Bertl those are scripts of util-vserver too 1150597762 M * Nam ah 1150597811 M * Nam hmm... maybe i'll leave it as the way I am using it 1150597812 M * Nam hehe 1150597838 M * Nam i'm using PHP/Bash to do it all manually, it works fine 1150597858 M * Bertl maybe that's why no slackware support is there yet :) 1150597859 M * Nam hehe, sounds like it's going to be a lot of work for me to get it working 1150597884 M * Nam possibly... maybe when i get some time i'll do it... for now, i have way to many other things going 1150597927 M * Nam btw... who came up with the css for the "Flower Page" hehe 1150597949 M * Nam hehe, you a bunch of stoners? 1150597961 M * Bertl which one? 1150597982 M * Nam i've only see one flower page 1150597994 M * Bertl which css stylesheet :) 1150597999 M * Nam it's the page that shows all the configure options 1150598007 M * Nam i've only seen one 1150598015 M * Nam i didn't know there was others 1150598017 M * Bertl it has at least 6 different ones 1150598041 M * Nam at the top of the page it says send them in 1150598049 M * Bertl your browser should allow you to select one 1150598069 M * Nam how do you select the different css themes? .... where in the page? 1150598089 M * Bertl not in the page, it's a selection like the magnification 1150598096 M * daniel_hozac View -> Use Style in mozilla browsers. 1150598106 M * Nam ah 1150598112 M * Nam i never knew about that 1150598117 M * Nam thanks daniel_hozac 1150598285 M * Nam the default one was the one i was referring to 1150598304 M * Nam a like purple/pink background with a pot leaf 1150598312 M * Nam tiles 1150598315 M * Nam tiled rather 1150598318 M * doener ah, .17 is out :) 1150598330 M * Bertl ah, yes 1150598343 M * Bertl means I have to update the patches :) 1150598376 M * Nam .17 of what? 1150598386 M * daniel_hozac the kernel, of course. 1150598393 M * doener .17 of 2.6 ;) 1150598415 M * Nam oh, right... i thought there was going to be a 2.6.16.21, hehe 1150598446 M * Nam i'm going to head to the store while this downloading is still gonna take at liest 15 minutes... bbiab 1150598458 M * Bertl k, cya 1150599641 Q * doener Quit: hgundadrd 1150599658 M * Nam k, upgrading now 1150599667 M * Bertl up? 1150599694 M * Nam hehe. right 1150599696 M * Nam downgrading 1150599714 M * Nam i do it using the upgradepkg command, that's why i typed up 1150599914 M * Nam recompiled and re-installed dietlibc 1150599935 M * Nam remaking util-vserver 1150600331 M * Nam done recompile of util-vserver and re-installed, recompiling kernel now 1150600365 M * Bertl check the tools , I doubt the compiler will mess up the kernel 1150600637 M * Nam vserver foo start 1150600637 M * Nam /lib/util-vserver/functions: line 473: 12567 Segmentatvserver foo start 1150600637 M * Nam /lib/util-vserver/functions: line 473: 12567 Segmentation fault $_VSERVER_INFO - FEATURE namespaceion fault $_VSERVER_INFO - FEATURE namespace 1150600663 M * Bertl hmm, that's interesting ... 1150600669 M * daniel_hozac very. 1150600689 M * Nam kernel's still compiling 1150600692 M * Bertl is there something 'hardened' in slackware? 1150600718 M * Bertl i.e. some special security modifications to gcc? 1150600724 M * Nam well.. this is kernel level, right, linux is linux, it's just how it's put together, and we are changing the kernel 1150600735 M * Nam this was working until i upgraded 1150600742 M * Nam good thing this is a test system 1150600742 M * Nam hehe 1150600777 M * Bertl well, I'm kind of worried about a distro which basically 'kills' userspace tools :) 1150600845 M * Nam hehe, it's not the distro 1150600864 M * Bertl well, I'm pretty sure we will figure _what_ it is sooner or later 1150600872 M * Nam all a distro is, is a bunch of software, the linux kernel, and then how it's all packaged 1150600884 M * Nam oh,... hehe, I WILL find this out 1150600890 M * Nam with help of course 1150600895 M * Nam but it need to have it working 1150600921 M * Nam once i recompile this kernel, and activate it, see how things go then 1150600949 M * Bertl okay 1150601124 M * Nam rebooting with new kernel now 1150601502 M * Nam damn it... somethings messed up now 1150601558 M * Nam i'm going to have to downgrade the entire system, i bet it's the glibc causing the problem 1150602647 J * tek_ ~tek@83.203.223.235 1150602726 M * Bertl welcome tek_! 1150603071 Q * PilatomiK Ping timeout: 480 seconds 1150603324 Q * Aiken Ping timeout: 480 seconds 1150605554 Q * derjohn2 Ping timeout: 480 seconds 1150605566 J * derjohn2 ~aj@dslb-084-058-204-018.pools.arcor-ip.net 1150606750 Q * tek_ Quit: Quitte 1150607555 J * f_ ~f_@83-215-237-1.seek.stat.salzburg-online.at 1150607707 Q * f_ Quit: 1150609195 T * Bertl http://linux-vserver.org/ | latest stable 2.01, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.{0.2,1.1}-rc23 | util-vserver-0.30.210 | libvserver-1.0.2 & vserver-utils-1.0.3 | He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;) 1150609217 M * Bertl rc23 for 2.6.17, enjoy :) 1150609271 M * Bertl I'm off to bed now ... have a good one everyone .. cya later ... 1150609280 N * Bertl Bertl_zZ 1150610151 Q * Greek0 Read error: Connection reset by peer 1150611087 Q * locksy Ping timeout: 480 seconds 1150612575 J * dna ~naucki@dialer-169-129.kielnet.net 1150612714 Q * dna Quit: 1150614100 J * meandtheshell ~markus@85-124-232-114.work.xdsl-line.inode.at 1150615725 J * cmatei ~cmatei@home-307461.b.astral.ro 1150616091 M * Nam well... Bertl, I know your offline, but hopefully to see this... I the system back to 10.2 release of the distro, and recompiled everything, it's working fine now, so it must be something with the new compiler version 1150617156 J * zkbrsnie ~zkbrsnie@83-64-146-226.klosterneuburg.xdsl-line.inode.at 1150617866 J * locksy ~locksy@mrtg.sisgroup.com.au 1150617917 J * _coocoon_ ~coocoon@84.160.112.198 1150617924 M * _coocoon_ morning 1150618584 J * pflanze ~chris@84-73-58-191.dclient.hispeed.ch 1150618884 M * Nam mornin 1150618890 M * Nam still night here 1150618893 M * _coocoon_ morning 1150620075 J * f_ ~f_@83-215-237-1.seek.stat.salzburg-online.at 1150620291 Q * bubulak Quit: Lost terminal 1150623061 M * gdm hi, i have a question about ssh 1150623506 A * pflanze waits for the question to appear.. 1150623525 M * pflanze just ask (tm) :) 1150623713 M * gdm pflanze: yeah, i know.. it was on behalf of someone else, but then he wasn't too bothered... 1150623740 M * pflanze ic 1150623746 M * gdm it seems that when he ssh's between the host and vserver, or vservers themselves, he can enter password but then it hangs 1150623764 M * gdm he says all the config settings are correct 1150623780 M * gdm 05:34 util-vserver 0.30.210-10 1150623780 M * gdm 05:35 linux-image-2. 2.6.16 vserver from backports 1150623781 M * gdm 05:35 install, and start/stop works brilliant 1150623792 M * gdm so i suggested he wrote to the list for more help 1150623799 M * gdm or joined the irc channel 1150623862 M * pflanze My first guess was DNS reverse lookup problems, but I think then it wouldn't get to the password entering stage. 1150623902 M * pflanze You could use strace and tcpdump. 1150623914 M * _coocoon_ gdm: how does he try to login as root 1150623954 A * gdm asking 1150624292 N * sukria_ sukria 1150624299 Q * DreamerC Ping timeout: 480 seconds 1150624324 M * gdm no response.. never mind ;-) 1150624452 M * _coocoon_ gdm: does it ever work? 1150624472 J * DreamerC ~dreamerc@59.112.26.30 1150624495 M * gdm he said earlier that vserver ... enter worked fine 1150624523 M * gdm he installed it on the laptop, but now is not answering questions anymore 1150624653 M * _coocoon_ gdm: he won't answer questions anymore? 1150624931 J * bonbons ~bonbons@83.222.39.166 1150625073 M * gdm i think he is focussing on other work, rather than experimenting with vservers.... 1150625101 M * gdm he says he tried to login as root 1150625127 Q * Hollow Quit: Konversation terminated! 1150625160 M * _coocoon_ gdm: i have asked that because of it is maybe set in the config that it isn't allowed to login as root 1150625289 J * Hollow ~hollow@home.xnull.de 1150625334 J * startx ~startx@p213.54.145.139.tisdip.tiscali.de 1150625389 M * gdm whoops! just realised i mistyped, _coocoon_ 1150625405 M * startx hi gdm 1150625406 M * gdm he says he tried to login as user - sorry! 1150625412 M * startx ;) 1150625421 M * gdm anyway, _coocoon_ - it is startx that is having the problems 1150625432 M * startx hi _coocoon_ 1150625438 M * _coocoon_ hi 1150625442 M * _coocoon_ startx 1150625456 M * startx i dunno how much gdm rold you allready 1150625482 M * _coocoon_ startx: i dunno if i can help u 1150625500 M * _coocoon_ startx: another idea is to chekc the ssh keys 1150625511 M * _coocoon_ startxmaybe delete them and try it again 1150625522 M * startx basically i created two vservers, set up is x.x.x.11 for host .13 and .14 are the vservers 1150625535 M * startx teh server run fine 1150625551 M * startx i can ping and nmap them, i entered them (via vserver command) 1150625556 M * startx and set up sshd 1150625564 M * startx when i do something like 1150625572 M * startx ssh -start x.x.x.13 1150625579 M * startx i get prompted for passwd 1150625585 M * startx i enetr passwd 1150625598 M * startx and then nothing 1150625622 M * startx ssh -l startx x.x.x.13 1150625623 M * _coocoon_ startx: r there error logs available 1150625627 M * Nam check the logs? 1150625660 M * Nam if your prompted for a password, you should log something within the vserver logs from ssh 1150625690 M * startx (pam_unix) session opened for user startx by (u 1150625690 M * startx d=0) 1150625690 M * startx Jun 18 08:21:34 bre sshd[1849]: error: openpty: No such file or directory 1150625690 M * startx Jun 18 08:21:34 bre sshd[1852]: error: session_pty_req: session 0 alloc failed 1150625693 M * startx Jun 18 08:21:40 bre sshd[1852]: (pam_unix) session closed for user startx 1150625733 M * bonbons startx: make sure you have /dev/ptmx inside your guests 1150625744 M * startx right 1150625754 M * bonbons also check wether /dev/pts gets mounted. 1150625804 M * bonbons if either is missing you should not be able to enter them via vserver login either (but vserver enter will work) 1150625822 M * startx ic 1150625863 M * startx thx, i'll check tat later, unfort. i booted into non vserver kernel now. 1150626033 M * startx brb 1150626035 Q * startx Quit: Lost terminal 1150626073 M * Nam i'm getting this error... although it doesn't seem to be effecting anything yet, it's happening when i run "vserver enter" 1150626074 M * Nam /dev/pts/0: Operation not permitted 1150626160 M * bonbons Nam: are you running vserver enter from a ssh connection? 1150626212 M * bonbons if so /dev/pts/0 is used by host and not available for guest -> which causes your error message. You should use vserver login in that case 1150626221 M * Nam yup 1150626230 M * Nam ah, ok 1150626232 M * Nam thx 1150626291 M * Nam when i run it from a console on the machine itself, i get this error... 1150626311 M * Nam /dev/vc/1: No such file or directory 1150626320 Q * _coocoon_ Read error: Connection reset by peer 1150626346 M * bonbons that's similar reason 1150626371 M * Nam that need to be created within the vserver? that file exists on the main system 1150626470 J * _coocoon_ ~coocoon@84.160.112.198 1150626505 M * bonbons do you have /dev/tty in your guest? 1150626523 M * Nam yup 1150626565 J * startx ~startx@p213.54.145.139.tisdip.tiscali.de 1150626615 M * bonbons isn't /dev/vc/* equivalent to /dev/tty*? 1150626644 M * nebuchadnezzar it is 1150626654 M * nebuchadnezzar is was the name of tty with devfs 1150626672 M * startx aha, interesting 1150626683 M * startx now it works 8) 1150626814 M * startx i recreated the user and now it worked fine 1150626855 M * startx (so maybe it was me who did some stupid thing before) 1150627004 M * gdm startx: probably ;-) 1150627364 M * Hollow bonbons: a few minutes i ago the first ipv6 packet found it's way to my box :) 1150627380 M * bonbons Hollow: fine :) 1150627404 M * bonbons currently I'm replacing my server's sysvinit by initng 1150627413 M * Hollow do you use tspc with freenet6? 1150627434 M * bonbons Jep, but a bit patched! it's quite some work to get all the initscripts right! 1150627479 M * bonbons I made it write it's logfile to /var/log/ and write a pid file. Then I also adjusted the template script to not touch my radvd config 1150627486 M * Hollow yeah, it didn't work with freenet6, i tried during verification at sixxs 1150627499 M * Hollow but with sixxs i can configure it via iproute and all is fine :) 1150627507 M * bonbons which version of tspc are you using? 1150627532 M * Hollow i tried both the in-portage one (total sucker) and the upstream one 1150627562 M * bonbons I have one in my overlay (tspc2) 1150627595 J * yarihm ~yarihm@80-218-2-149.dclient.hispeed.ch 1150627596 M * Hollow well, since i got approved at sixxs i don't need freenet6, and the PoP is quite fast 1150627597 M * bonbons but when using UDP tunnel you need to load tun <- easy to forget and will cause tunnel not to come up 1150627636 M * bonbons so vserver-utils homepage is on IPv6 now? 1150627638 M * Hollow try: ping6 2001:a60:f000:33::2 1150627643 Q * startx Quit: Lost terminal 1150627652 M * Hollow no, it's my local box here (home.xnull.de) 1150627685 M * bonbons 250ms ping from here 1150627720 M * Hollow not the best, but ok.. 1150627753 M * bonbons sufficient for using IPv6, but not for shooters ;) 1150627791 M * Hollow i don't play games ;) 1150627813 M * bonbons I don't play much either 1150627824 M * Hollow do you get full adsl speed with your pop? 1150627833 M * bonbons traceroute6 gave me 17 hops to you 1150627845 M * Hollow does the pop even influence the speed? 1150627849 M * Hollow (like a proxy) 1150627876 M * bonbons yep, but they don't offer speedy lines, I've just 1024 down, 128 up 1150627894 M * Hollow hm, seems i can use full adsl speed on the ipv6 tunnel 1150627932 M * bonbons here it's my ADSL that's slow, IPv6 gets as much as IPv4 (minus tunneling overhead) 1150627935 M * Hollow fine fine.. guess the pop is just 2-3km away from here :) 1150627941 M * Hollow ah, ok 1150628011 M * Hollow also the "replace NAT with ipv6" sounds veryx interesting 1150628046 M * bonbons once initng is working fine on my server it will get a IPv6 enabled vserver kernel. Just bad luck that IPv6 is not yet much in use... 1150628085 M * Hollow yeah, i always wondered .. time to sell native IPv6 option with adsl lines :) 1150628205 M * bonbons but as far as I can see ISPs are not on that track, they prefer selling static IPv4 addresses at high price 1150628222 M * Hollow can you try ping6 2001:a60:f000:33::3 ? 1150628268 M * bonbons and then all ADSL routers whould have to be upgraded, saw none with IPv6 support yet (unless with a non-original linux-firmware) 1150628337 M * bonbons no route to host 1150628345 M * Hollow ok, i somewhat expected that 1150628353 M * Hollow still have to wait for a whole subnet 1150628380 M * bonbons last hop is your broker (demuc02.sixxs.net) 1150628408 J * experriment ilia@195.62.22.86 1150628411 M * Hollow yeah, i have to get some credits (by keeping the tunnel online) to be able to request a subnet 1150628476 M * bonbons ok, that's the item that freenet6 does not ask for 1150628524 M * Hollow protection against abusal i guess 1150628623 M * bonbons possibly, or just a policy - they target other kind of users 1150628859 P * experriment 1150628968 M * bonbons and oftc is not IPv6 enabled it seems :( 1150629099 J * lilalinux ~plasma@dslb-084-058-204-105.pools.arcor-ip.net 1150629419 Q * Hollow Remote host closed the connection 1150629423 J * Hollow ~hollow@home.xnull.de 1150629438 M * Hollow irc6.oftc.net :) 1150629555 M * Hollow hm 1150629560 M * Hollow it fallback to ipv4 1150629686 Q * Hollow Quit: 1150629925 Q * zkbrsnie Read error: Connection reset by peer 1150629948 M * Wonka "host irc6.oftc.net" told me 6 v4 addresses and 3 v6 1150629979 J * Viper0482 ~Viper0482@p54977D6A.dip.t-dialin.net 1150630149 M * bonbons seen, but no IPv6 capable DNS server for the zone 1150630154 M * Nam how do you specify the hostname for the guest system using the new config style? 1150630213 M * mnemoc on uts/nodename 1150630221 J * Hollow ~hollow@cl-52.muc-02.de.sixxs.net 1150630240 M * Nam thx 1150630264 M * mnemoc the weed page is your friend :) 1150630268 M * mnemoc http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1150630288 M * Hollow ah it works 1150630299 M * bonbons Hollow -> Hollow6 1150630304 M * Hollow indeed 1150630305 M * Nam yea, i have it open on the other page, wanted to make sure it was that 1150630315 N * Hollow Hollow6 1150630329 M * Nam hehe, weed page.... speaking of which... *packs bowl* 1150630470 M * Nam is it true you can only have like 16 interfaces per vserver? 1150630479 M * Nam ie, 16 ips only 1150630494 M * bonbons Nam: unless you changed the define yes 1150630518 M * bonbons more liberty on address count will come in near future 1150630528 M * Nam is there any way to add ip's without rebooting the vserver? 1150630564 M * bonbons yep, but I don't know the util-vserver command for that 1150630735 M * Nam i noticed there is a -s/--sync option, don't quiet knot how to use it though 1150630737 M * _coocoon_ Nam: here it is point 8 http://deb.riseup.net/vserver/usage/ 1150630762 M * Nam perfect 1150630765 M * Nam thanks _coocoon_ 1150630771 M * _coocoon_ uw 1150631128 N * gdm effraie 1150631154 N * effraie gdm 1150631821 J * Greek0 ~greek0@85.255.145.201 1150632166 Q * anonc helium.oftc.net xenon.oftc.net 1150632166 Q * _Kara helium.oftc.net xenon.oftc.net 1150632166 Q * starlein helium.oftc.net xenon.oftc.net 1150632166 Q * jake- helium.oftc.net xenon.oftc.net 1150632166 Q * glen helium.oftc.net xenon.oftc.net 1150632166 Q * ag- helium.oftc.net xenon.oftc.net 1150632166 Q * morrigan helium.oftc.net xenon.oftc.net 1150632256 J * morrigan morrigan@212.16.62.52 1150632256 J * glen ~glen@elves.delfi.ee 1150632256 J * ag- ag@caladan.roxor.cx 1150632256 J * jake- psybnc@murlocs.org 1150632256 J * starlein star@fo0bar.de 1150632256 J * _Kara ~Kashira@wan14.ihredomainadresse.de 1150632256 J * anonc ~anonc@staffnet.internode.com.au 1150632459 Q * s0undt3ch Ping timeout: 480 seconds 1150632719 J * s0undt3ch ~s0undt3ch@bl7-244-107.dsl.telepac.pt 1150633127 J * Wenix_ ~wenix@81.7.189.11 1150633128 Q * Wenix_ Quit: 1150633188 M * Hollow6 Bertl_zZ: works great! http://paste.linux-vserver.org/98 1150634372 M * derjohn is rc13 the lastest what is avail for 2.6.15? I want to patch a ubuntu*-server kernel (which is a 2.6.15) but rc13 seems a little old ... 1150634393 M * derjohn sladen, ping ? 1150635122 M * derjohn hm, 2.6.17 is out. maybe i should simply drop the ubuntu sources ... 1150635225 M * bonbons derjohn: I guess some work to check that latest patches work on brand new 2.6.17 ;) 1150635247 M * derjohn i press thumbs, that i will appy cleanly ... . 1150635309 M * bonbons bus probably rc23 came out at about same time 1150635336 M * derjohn ahhhhhhhhh yeah. i just saw it .... 1150635458 N * Bertl_zZ Bertl 1150635471 M * Bertl morning folks! 1150635489 M * bonbons morning Bertl! 1150635492 M * derjohn bonbons, applies cleanly :) 1150635512 M * derjohn Bertl, moin + thanks for the quick rc23 ;) 1150635536 M * bonbons derjohn: an I will have to check my IPv6 on .17 + rc23 1150635536 M * _coocoon_ hello bertl have powered off the machine 1150635584 M * Bertl derjohn: you're welcome! but you have to thank doener for spotting the kernel release, I would have missed in the evening :) 1150635592 M * derjohn bonbons, status of the v6 patch? 1150635597 M * Bertl _coocoon_: that's fine 1150635630 M * bonbons TCP+UDP on unicast works fine 1150635641 M * derjohn Bertl, next time I send you an SMS when a new kernel is released if you like :) 1150635653 M * bonbons just avoid using sockets that do IPv4 and IPv6 at same time as it may cause trouble with socket conflicts 1150635668 M * derjohn bonbons, and it doesnt brake other things? so: complining in does not harm? 1150635706 M * bonbons I saw no issue yet, so it shouldn't harm 1150635808 M * bonbons for guests IPv6 is denied if no such address is associated to the guest, for the rest, rules are similar as for IPv4 1150635945 Q * mire Quit: Leaving 1150635995 M * Hollow6 rc23 works http://paste.linux-vserver.org/100 1150636003 M * Hollow6 Bertl: and reset too 1150636030 M * Bertl and v6? 1150636052 M * Hollow6 works, but i didn't test the vserver v6 patch yet 1150636069 M * Hollow6 ping6 home.xnull.de 1150636071 M * Bertl just wondered about the Hollow6 :) 1150636078 M * Hollow6 :p 1150636113 M * Hollow6 will test it soon 1150636134 M * derjohn in 2.6.17 'Multi-core scheduler support' <-- any issules with the vserver scheduler? 1150636138 M * derjohn *issues ? 1150636138 M * Bertl daniel_hozac: could you give your magic scripts a run (if not done so already) on the 2.6.17 patches? 1150636154 M * Bertl derjohn: not that I would know of ... 1150636182 M * Hollow6 Bertl: would this seem reasonable? http://home.xnull.de/misc/limits.png 1150636230 M * Bertl time is seconds or minutes? 1150636236 M * Hollow6 minutes 1150636236 M * Bertl or hours? 1150636260 M * Bertl processes is graphed? 1150636284 M * Hollow6 i first thought of cpu usage 1150636318 M * Hollow6 but it may coincide with processes ;) 1150636352 M * Bertl but looks good and nice :) 1150636367 M * Hollow6 you know i like to visualize things ;) 1150636392 M * Bertl yeah, just missed the legend and units :) 1150636444 M * Hollow6 yeah, but it should at least be valid for all limits, if it's reasonable is another question.. 1150636462 M * Hollow6 i.e. cur always between min/max 1150636610 M * Bertl I think, now that iattrs are in jfs/mainline, we should make the jfs changes stable 1150636617 M * Bertl any objections? 1150636880 M * Bertl afk, breakfast, brb 1150637936 M * Bertl back now 1150638065 M * daniel_hozac was vs_pid.h moved to sched.h? 1150638250 M * Bertl hmm, we had that patch for testing, but IIRC we decided to leave it out (might have gotten in though) 1150638272 M * daniel_hozac it appears to be in -rc23. 1150638299 M * Bertl okay, guess we revert it then 1150638625 M * daniel_hozac http://paste.linux-vserver.org/101 1150638658 M * daniel_hozac i already removed the ones i know are bogus. 1150638709 M * Bertl ah, cool 1150638795 M * Bertl I'll deal with the cvirt, could you have a look at the locked_vm? 1150639043 M * daniel_hozac sure. 1150639831 J * vrwttnmtu ~eryktyktu@82-69-161-137.dsl.in-addr.zen.co.uk 1150639919 M * Bertl welcome vrwttnmtu! 1150639944 M * vrwttnmtu Hello, ever cheerful Bertl 1150639953 M * vrwttnmtu (Sure it's a bot) 1150639993 M * vrwttnmtu Is your ADSL lag problem better now? 1150640025 A * Hollow6 giggles 1150640029 M * Bertl well, I _wish_ it was ADSL (it is dialup :/) but it is much better now 1150640033 N * Hollow6 Hollow 1150640049 M * vrwttnmtu Dialup? *shudders* 1150640056 M * vrwttnmtu Hello Hollow. 1150640076 M * vrwttnmtu I see Hollow is v6ing 1150640087 M * Hollow yup 1150640105 M * phedny Hollow: woei, sixxs :) 1150640115 M * vrwttnmtu I'm v6'd here, but neutron.oftc.net obviously doesn't return aaaa records. 1150640142 M * Hollow phedny: yeah, awesome 1150640155 M * Bertl well, not only the a* records cause issues as it seems 1150640167 M * Hollow i used irc6.oftc.net 1150640177 M * vrwttnmtu Let me try that 1150640178 M * vrwttnmtu BRB 1150640181 Q * vrwttnmtu Quit: Leaving 1150640198 M * mnemoc having a and aaaa, are all used on the round-robin? 1150640203 J * vrwttnmtu ~eryktyktu@2001:618:400:e563::1 1150640205 M * phedny mnemoc: yes 1150640209 M * vrwttnmtu Smell the same 1150640211 M * vrwttnmtu Feels the same 1150640212 M * Hollow muhe 1150640232 M * vrwttnmtu Got my /64 at home, and my /48 at work 1150640234 M * Hollow vrwttnmtu: but ~eryktyktu@2001:618:400:e563::1 1150640242 M * mnemoc vrwttnmtu: but now you don't have a reverse :) 1150640242 M * vrwttnmtu Indeedy 1150640251 M * vrwttnmtu I get it from a tunnel-broker 1150640254 M * phedny mnemoc: a node requests for example AAAA record of a hostname and it gets a list of IPv6 addresses returned 1150640260 M * phedny in random order 1150640263 M * vrwttnmtu I thought they did reverse DNS 1150640294 M * phedny vrwttnmtu: with SixXS you must run DNS yourself, they only provide the hookup point (NS record) 1150640296 M * daniel_hozac Bertl: http://daniel.hozac.com/vserver/delta-infiniband-fix02.diff 1150640320 M * vrwttnmtu Hollow, It's amazing how many people have v4 firewall rules, and they just allow v6 to rampage around their networks. 1150640355 M * vrwttnmtu Bertl, Hollow, it's actually v6 that I came in to ask about. I'd really really love it. How close is it to being supported? 1150640360 M * phedny vrwttnmtu: mostly caused by unawareness (or ignorance in my case :p) 1150640380 M * vrwttnmtu phedny, I use the BTExact tunnel broker 1150640396 M * vrwttnmtu I'm going to log in, and see why they're not supplying reverse 1150640462 M * sladen derjohn: yo 1150640490 M * bonbons vrwttnmtu: well, there's a patch of mine that offers IPv6 for guests, UDP+TCP+Ping6 1150640528 M * derjohn sladen, just wanted to know if you did patch the dapper kernel with some current vserver patch? 1150640529 M * vrwttnmtu bonbons, How well does it work? Is it stable? 1150640529 M * bonbons but it didn't get tested that much yet and still has to be ported to 2.6.17 ;) 1150640539 M * Hollow bonbons: do the patches work with 2.6.17? 1150640565 M * bonbons Hollow: not looked at it yet, but it's on my todo 1150640570 M * vrwttnmtu Aaah. I don't mind running a not-the-latest-kernel-ever (as long as there aren't any security bugs in that version) to get a chance to run v6 guests 1150640594 M * daniel_hozac Bertl: -rc23 seems to lack delta-peer-fxi01? 1150640612 M * vrwttnmtu bonbons, If you were running a few vservers at work, for RADIUS, Nagios, etc, would you trust your patch with it? :) 1150640614 M * bonbons vrwttnmtu: it's beta 1150640672 M * bonbons vrwttnmtu: I wouldn't put it on production server yet, lacks testing 1150640696 M * vrwttnmtu bonbons, OK, thanks for the advice. Can you send me a link though, and I'll try it out at home for sure. 1150640725 M * bonbons http://homepage.internet.lu/brunop/vserver/ 1150640783 M * bonbons take the latest (currently patch-2.6.16.20-vs2.1.1rc22-ipv6e.diff) 1150640827 M * vrwttnmtu OK, let me get patching. 1150640828 M * vrwttnmtu :) 1150640843 M * phedny bonbons: couldn't your IPv6 patch be integrated with the vserver development version? 1150640929 M * bonbons will be in nore or less near, need to check with Bertl when -- either before address matching changes or afterwards 1150640964 M * daniel_hozac along with lo. 1150641035 M * vrwttnmtu how can I find the location to a specific vserver patch version ( vs2.1.1rc22 ) ? 1150641040 M * derjohn sladen, /me away now.... if you have a patch already, let me know. thx. 1150641074 M * daniel_hozac vrwttnmtu: http://vserver.13thfloor.at/Experimental/ 1150641100 M * vrwttnmtu Merci... 1150641182 A * vrwttnmtu can't wait for my Ipv6 guests :) 1150641191 A * vrwttnmtu can't wait for **his** IPv6 guests. 1150641192 M * vrwttnmtu :) 1150641215 A * phedny still needs to find an affordable colocation provider that supports IPv6 :p 1150641233 M * sladen derjohn: no. Not even thought about it 1150641283 M * Bertl daniel_hozac: it should be in devel, IIRC 1150641462 M * Bertl nah, isn't :) 1150641486 M * Bertl do we consider it suitable for stable too? 1150641522 M * Wonka how far is NGNET? 1150641665 M * Bertl very far atm :) 1150641725 Q * m4z Quit: brb 1150641736 J * m4z m4z@bastard-operator.from-hell.net 1150641891 M * daniel_hozac Bertl: i think it should be in stabΓƒle too, given that it breaks the hiding for (and possibly using, i don't remember) P-t-P addresses. 1150642066 M * Bertl okay, good argument, we did have some testing too IIRC 1150642067 M * bonbons Bertl: should I keep doing an incremental patch of ipv6 + modular? If I write some core do implement address masks/ranges for guests should that be pre-IPv6 or post-IPv6? 1150642097 M * Bertl bonbons: keep it modular if possible 1150642130 M * bonbons it's just a matter of knowing how many cowed kernel trees to setup 1150642150 M * daniel_hozac a bazillion! :) 1150642150 M * Bertl the smaller the changes, the sooner they'll get in 1150642164 M * bonbons so base vs tree; + IPv6; + IPv6mod 1150642261 M * vrwttnmtu Japan 0 - Croatia 0 1150642269 M * mnemoc finished? 1150642273 M * vrwttnmtu Yep 1150642289 M * Bertl I think we have some vserver users in Japan :) 1150642301 M * vrwttnmtu And probably some in .hr :) 1150642304 M * Bertl vrwttnmtu: so your stats must be wrong .) 1150642310 M * vrwttnmtu Heh :) 1150642315 M * bonbons will do that for now... and see how I do for experimenting with new address matching 1150642341 M * vrwttnmtu Do any VS technologies do IPv6 yet? If not, it would be great if vserver got the jump on them 1150642386 M * daniel_hozac didn't OpenVZ just add IPv6 support? 1150642436 M * Bertl probably kir knows ... 1150642447 M * bonbons that's also what I thought, and things like UML do also 1150642465 M * mnemoc has anyone played with vserver and gfs2 for clustering? 1150642481 M * Bertl ocfs2 is _almost_ supported 1150642971 M * Bertl guess we might drop the SPLIT stuff completely now 1150643036 M * Bertl 8after all, mainline caught up with that :) 1150643150 A * Wonka wants iptables in vserver guests :( 1150643173 M * Bertl Wonka: for what purpose? 1150643184 M * Wonka Bertl: Users want firewalling. 1150643218 M * Bertl daniel_hozac: could you check why we want/need vs_cvirt/vs_pid in arch/um/{drivers/mconsole_kern.c,kernel/process_kern.c} please? 1150643235 M * Bertl Wonka: why not give them firewalling (on the host)? 1150643238 M * Wonka "real" network stack virtualization would be nice too. would bring, afaics, IPv6 (and every other network protocol), iptables, own lo for every guest, ..., all at once... 1150643244 M * Wonka Bertl: how? 1150643263 M * Bertl make an iptables chain for every guest 1150643265 M * Wonka i don't want to give iptables access to them. i don't really trust this special user. 1150643280 M * Bertl allow them to add their rules via web interface 1150643281 M * Wonka mmh. would be a possibility. 1150643312 M * Wonka the very same user bitches about not being able to use raw sockets 1150643323 M * Wonka cause it disables traceroute, apparently 1150643339 M * Wonka (yes, user=id10t) 1150643340 M * phedny vrwttnmtu: any hardware-emulating products (like VMWare, Virtual PC) have IPv6 support, since nothing special is needed for them to make it work ;) 1150643348 M * Bertl Wonka: rm -f traceroute && ln tracepath traceroute 1150643432 M * Wonka Bertl: gone after next upgrade. and also, he'll find another thing to bitch about. 1150643460 M * Wonka *sigh* if he would just leave, as he threatens to do every second time he talks to me... 1150643476 M * Bertl Wonka: yeah, he wants to sniff other guests and fake packets, are you going to provide that 'feature'? :) 1150643495 M * vrwttnmtu bonbons, Why is menuconfig telling me that ip6tables support is obselete? 1150643517 M * bonbons what mainline kernel? 1150643517 M * Wonka Bertl: surely not. that's why i want NGNET, and only route _his_ stuff to him. 1150643531 M * vrwttnmtu 2.6.20.16 1150643535 M * vrwttnmtu + vserver 1150643538 M * vrwttnmtu + ipv6 patch 1150643588 M * Bertl Wonka: well, that will not work, unless you really _do_ routing (means twice the overhead) 1150643604 M * bonbons vrwttnmtu: hm, don't remeber seeing that, where is it in menuconfig? 1150643613 M * vrwttnmtu β”‚ β”‚ [*] Network packet filtering (replaces ipchains) ---> β”‚ β”‚ 1150643621 M * vrwttnmtu β”‚ β”‚ IPv6: Netfilter Configuration (EXPERIMENTAL) ---> β”‚ β”‚ 1150643628 M * vrwttnmtu β”‚ β”‚ < > IP6 Userspace queueing via NETLINK (OBSOLETE) β”‚ β”‚ 1150643636 M * vrwttnmtu Which is the only option in there 1150643651 M * Bertl Wonka: for example, take the raw sockets, you basically have to decide what you want here: a) he should be able to send arbitrary packets, b) he onkly can send ip packets from/to his range 1150643721 M * Wonka Bertl: mmh. maybe take something like uml_switch? 1150643722 M * bonbons vrwttnmtu: I have just [*] Network packet filtering (replaces ipchains) ---> 1150643732 M * Bertl Wonka: if I buy a dedicated server from you, and start arp spoofing to reroute your router traffic, is that fine with you? 1150643741 M * vrwttnmtu bonbons, And, inside that option? 1150643745 M * Wonka Bertl: surely not. 1150643770 M * Bertl Wonka: so why isn't it okay to block/prevent that in the first place then? 1150643801 M * Wonka Bertl: but if i buy a dedicated server, i can send whatever my (self installed) OS let's me 1150643820 M * bonbons inside I have 3: [ ] Network packet filtering debugging, Core Netfilter Configuration -->, IP: Netfilter Configuration -->, IPv6: Netfilter Configuration (exp) --> 1150643822 M * Bertl Wonka: okay, so it would be fine to do that then? 1150643834 M * vrwttnmtu I have 5 1150643871 M * Wonka Bertl: the hosting provider would need to filter bad stuff anyway. why shouldn't vserver try to get quite near to that? 1150643900 M * vrwttnmtu Network packet filtering debugging, Bridged IP/ARP packets filtering, Core Netfilter Configuration, IP: Netfilter Configuration, IPv6: Netfilter Configuration (EXPERIMENTAL), Bridge: Netfilter Configuration 1150643909 M * Bertl Wonka: because it doesn't make sense to allow 'evil' packets and then filter them lateron, consuming resources for nothing ... 1150643919 M * trippeh Bertl: I'm getting a syntax error with the 2.6.17-2.0.2-rc23 patch 1150643922 M * Bertl Wonka: do you filter arp spoofs? 1150643940 M * Bertl trippeh: please upload to paste.linux-vserver.org 1150643940 M * bonbons vrwttnmtu: what filtering did you select? 1150643952 M * vrwttnmtu bonbons, Filtering? Where? 1150643974 M * Wonka Bertl: not that i _want_ to filter out bad ARP. but having proven old tools like traceroute fail is also somehow not nice. 1150643979 M * trippeh Bertl: http://paste.linux-vserver.org/102 1150644001 M * Wonka Bertl: at the moment not, because the roots in my club are trustworthy. 1150644009 M * Bertl Wonka: that's something different 1150644028 M * Bertl (the traceroute) 1150644034 M * vrwttnmtu Wonka, There's always tracepath 1150644036 M * bonbons vrwttnmtu: if you want IPv6 stateful filtering you must first disable the connection tracking under IPv4, then choose "Layer 3 Independent Connection tracking" in core 1150644043 M * vrwttnmtu I think that doesn't need raw sockets 1150644057 M * Bertl trippeh: will be fixed shortly 1150644092 M * trippeh Bertl: Ok, thanks 1150644096 M * Wonka vrwttnmtu: $LUSER does know traceroute, doesn't know tracepath, believes he was told he was "full root", and gets "traceroute: raw socket: Operation not permitted" 1150644113 M * vrwttnmtu Wonka, $LUSER needs re-educating. 1150644114 M * vrwttnmtu :) 1150644125 M * vrwttnmtu Or $LUSER needs UML 1150644132 M * Wonka vrwttnmtu: everyone who knows him and you knows that... 1150644137 M * bonbons vrwttnmtu: for me IP6tables is not marked as obsolete, just the netlink above 1150644208 M * bonbons s/netlink/userspace queuing/ that's because there is new netlink and netfilter puts more and more IPv4/IPv6 code in common 1150644208 M * Bertl Wonka: given that there is somebody willing to test it, we could consider allowing certain raw packets (like used by traceroute) 1150644250 M * vrwttnmtu bonbons, Well, I'll give it whirl. I don't see the options for REJECT, LOG, ULOG, etc though. I'll compile it, and see what I have. 1150644264 M * Bertl vrwttnmtu: xtables? 1150644270 M * vrwttnmtu Wonka, alias traceroute=tracepath 1150644272 M * vrwttnmtu :) 1150644309 M * vrwttnmtu Bertl, I saw that - when did all this change? :) I'm running 2.6.15, and it's all "normal" in there. /grumpy old man: It wasn't like this in my days... :) 1150644325 M * daniel_hozac Bertl: find_task_by_pid? 1150644357 M * bonbons vrwttnmtu: that changed during the last few kernels, all started in january when IPv6 stateful filtering came in 1150644358 M * Bertl well, the differences between .15 and .17 are two magnitudes larger than the entire vserver patch :) 1150644374 M * Wonka *g* 1150644393 M * Bertl daniel_hozac: explains the vs_pid.h, but not the vs_cvirt, no? 1150644404 M * vrwttnmtu Damn this release early, release often :) 1150644409 M * vrwttnmtu I'm moving to BSD. :) 1150644419 M * daniel_hozac Bertl: which file? which branch? 1150644442 M * daniel_hozac Bertl: looks like the hunk for arch/um/kernel/process_kern.c in 2.0.2-rc23 is wrong too, it uses vs_pid.h. 1150644460 A * Wonka looks what traceroute _needs_ raw packets for... 1150644486 M * Bertl daniel_hozac: yeah, that is from stable 1150644512 M * mnemoc has anyone get tcptraceroute running from within a guesT? 1150644519 M * trippeh vrwttnmtu: At least they kept the userspace interface intact ;-) 1150644524 M * Bertl daniel_hozac: so that should be vs_cvirt.h then 1150644555 M * daniel_hozac Bertl: i don't see anything in that file requiring either. 1150644558 M * Bertl mnemoc: never tried, but tracepath works fine 1150644605 M * vrwttnmtu trippeh, Yeah, guess we should be thankful for small mercies :) 1150644614 M * Bertl daniel_hozac: okay, could you do a check-run for such unused includes too? 1150644661 M * daniel_hozac Bertl: well, the script checks for unused includes already. 1150644676 M * Bertl hmm .. okay, why did it miss those? 1150644679 M * daniel_hozac but it doesn't catch files from the other branch :) 1150644690 M * Bertl ah, okay, good, np 1150644696 M * daniel_hozac it checks if the tree has vs_pid and then enables all those checks. 1150644716 M * Bertl okay, leave it as is, will fix the vs_pid stuff first 1150644721 M * daniel_hozac there are a lot of vs_pid references in 2.0.2. 1150644745 M * Bertl yep, seems they slipped through 1150644753 M * vrwttnmtu Can't believe Budweiser is the sponsor of the World Cup. 1150645117 M * Wonka can't believe they're allowed to forbid people wearing clothes with other beer logos. 1150645138 M * Wonka "kann doch nicht warstein!" 1150645294 M * vrwttnmtu OK. Going for a reboot. 1150645295 M * vrwttnmtu :) 1150645296 Q * vrwttnmtu Quit: Leaving 1150645623 J * vrwttnmtu ~eryktyktu@82-69-161-137.dsl.in-addr.zen.co.uk 1150645630 M * vrwttnmtu Well, it's booted. 1150645726 M * vrwttnmtu So, bonbons, I can just add an interfaces/0/ipv6 file ? 1150645734 M * daniel_hozac not yet. 1150645745 M * vrwttnmtu not yet @ me? 1150645754 M * bonbons yep 1150645756 M * daniel_hozac as you're the only one talking :) 1150645775 M * vrwttnmtu I didn't know if you were answering to a conversation before I arrived 1150645788 M * vrwttnmtu daniel_hozac: not yet. bonbons: yep 1150645791 M * vrwttnmtu What's the answer :) 1150645807 M * daniel_hozac i guess the yep was for "not yet @ me" :) 1150645849 M * vrwttnmtu Aaah, OK. So how do I get the v6 address inside the guest? 1150645886 M * Bertl daniel_hozac: reg. infiniband, I'm not happy with the first hunk, as it basically disables the normal rlimit check ... 1150645902 M * Bertl daniel_hozac: the rest looks fine 1150645934 M * daniel_hozac i was unsure about the first hunk. 1150645966 M * bonbons vrwttnmtu: for now you have to manually add the IPv6 addresses using either vserver-utils's vncontext or use danizl_hozac's help tool for that 1150646008 M * vrwttnmtu -su: vncontext: command not found 1150646024 M * vrwttnmtu sys-cluster/util-vserver-0.30.210-r13 * 1150646035 M * bonbons I said vserver-utils's vncontext (vserver-utils != util-vserver !!) 1150646063 M * bonbons daniel_hozac: where is your tiny IPv6-addr adder? 1150646068 M * daniel_hozac http://daniel.hozac.com/vserver/chbind6.c 1150646180 M * daniel_hozac gcc -lvserver -I/include -ochbind6 chbind6.c IIRC. 1150646198 M * vrwttnmtu Just about to install vserver-utils from portage - do I need diet, or not? 1150646217 M * bonbons vrwttnmtu: that one is too old 1150646227 M * vrwttnmtu daniel_hozac, That worked 1150646246 M * vrwttnmtu What's the syntax for chbind6 ? 1150646255 M * daniel_hozac then chbind6 -n -a -6 1150646295 M * vrwttnmtu Sweet 1150646295 M * vrwttnmtu :) 1150646327 M * vrwttnmtu I see the host v6 addr in the guest though 1150646333 M * vrwttnmtu Sure I put a different one in there 1150646349 M * vrwttnmtu Yes, I did 1150646352 M * vrwttnmtu Hmm. 1150646353 M * bonbons did you set the right flags for you guest? 1150646361 M * vrwttnmtu bonbons, I don't know. 1150646362 M * vrwttnmtu :) 1150646387 M * vrwttnmtu I just use the same flags I normally use 1150646402 M * vrwttnmtu CAP_SYS_ADMIN 1150646405 M * vrwttnmtu And all of those 1150646407 M * vrwttnmtu (Not really) 1150646409 M * vrwttnmtu :) 1150646417 M * bonbons then you should also see the host's IPv4 addresses ;) 1150646431 M * vrwttnmtu bonbons, I see the guest's v4 address 1150646438 M * vrwttnmtu And the host's v6 address 1150646443 M * vrwttnmtu Muy extranjero 1150646480 M * vrwttnmtu And pinging the new v6 address just gives me Address unreachable 1150646493 M * daniel_hozac did you create the guest's address? 1150646505 M * daniel_hozac you have to set up the addresses on the host. 1150646507 M * vrwttnmtu Create it? Where? 1150646512 M * vrwttnmtu Aaah 1150646516 M * bonbons Then add VXF_HIDE_NETIF flag to your guest 1150646556 M * vrwttnmtu # vserver mx restart 1150646556 M * vrwttnmtu Unknown flag 'VXF_HIDE_NETIF' 1150646557 M * bonbons (it's a context capability [CCAP]) 1150646561 M * vrwttnmtu Aah 1150646561 M * vrwttnmtu :) 1150646578 M * daniel_hozac really? those are VXC_, no? 1150646583 M * daniel_hozac hide 1150646597 M * daniel_hozac ugh, laptop keyboards... hide_netif is the flag, but it's enabled by default for the tools. 1150646624 M * bonbons you're right daniel_hozac: it's bug in front of my keyboard 1150646661 M * vrwttnmtu So, do I add the guest v6 address to the main i/f on the host? 1150646674 M * vrwttnmtu ifconfig eth0 add 2001:124:ccccccc::1/64 ? 1150646705 M * bonbons vrwttnmtu: that doesn't really matter 1150646726 M * vrwttnmtu Ok. I've booted the vserver with that ccap added 1150646734 M * vrwttnmtu What's the next step 1150646746 M * bonbons just put it on the device you prefer (best is interface that is on the matching subnet unless you have IPv6 routing enabled on your box) 1150646771 M * vrwttnmtu It's going to be on the same /64 as the host 1150646780 M * vrwttnmtu But I have v6 routing enabled too for other things 1150646826 M * vrwttnmtu bonbons, You meant it doesn't matter which i/f I add it to? 1150646831 M * bonbons anyhow, for box-local testing it will work anyhow 1150646838 M * bonbons yep 1150646855 M * vrwttnmtu ifconfig eth0 add 2001:618:400:xxxx::25/64 1150646874 M * vrwttnmtu ./chbind6 -n 25 -a -6 2001:618:400:xxxx::25/64 1150646893 M * vrwttnmtu I see both the host and guest v6 address in the host now 1150646904 M * vrwttnmtu # ping6 www.kame.net 1150646904 M * vrwttnmtu PING www.kame.net(orange.kame.net) 56 data bytes 1150646904 M * vrwttnmtu 64 bytes from orange.kame.net: icmp_seq=1 ttl=54 time=928 ms 1150646904 M * vrwttnmtu 64 bytes from orange.kame.net: icmp_seq=2 ttl=54 time=812 ms 1150646904 M * vrwttnmtu --- www.kame.net ping statistics --- 1150646905 M * vrwttnmtu 3 packets transmitted, 2 received, 33% packet loss, time 1998ms 1150646907 M * vrwttnmtu rtt min/avg/max/mdev = 812.730/870.565/928.400/57.835 ms 1150646909 M * vrwttnmtu But I am connected 1150646910 M * bonbons that is correct 1150646911 M * vrwttnmtu Which is great 1150646923 M * vrwttnmtu Why do I see the host v6, but not the host v4 then? 1150646961 M * bonbons you just said you saw both addresses in the host 1150646970 M * vrwttnmtu Sorry 1150646974 M * vrwttnmtu I meant in the guest 1150646994 M * vrwttnmtu I see both v6 addresses in both host and guest 1150647494 M * bonbons I just get to see the right ones... 1150647614 M * bonbons vrwttnmtu: do the guest's addresses show up correctly in /proc/virtnet//info? 1150647654 M * Bertl trippeh, daniel_hozac: http://vserver.13thfloor.at/Experimental/patch-2.6.17-vs2.0.2-rc23.1.diff for testing 1150647678 M * vrwttnmtu bonbons, Yes, only the guest IP's are in there 1150647698 M * bonbons with what tool do you look at the addresses? 1150647707 M * daniel_hozac Bertl: hmm, still vs_pid references. 1150647709 M * vrwttnmtu ifconfig 1150647721 M * vrwttnmtu Haven't installed iproute2 yet 1150647726 M * Bertl daniel_hozac: ah, yes, I have a patch here, but didn't add it :) 1150647732 M * bonbons wha version? 1150647738 M * vrwttnmtu But it shouldn't display addresses that "don't exist" to it, should it? 1150647752 M * vrwttnmtu # ifconfig -V 1150647752 M * vrwttnmtu net-tools 1.60 1150647752 M * vrwttnmtu ifconfig 1.42 (2001-04-13) 1150647760 M * vrwttnmtu Nice and new 1150647760 M * vrwttnmtu :) 1150647771 M * bonbons it shouldn't! I have same version here... 1150647784 M * vrwttnmtu I'm amd64 if that's any difference 1150647794 A * vrwttnmtu strokes his lovely amd64 1150647812 M * bonbons ok, ip lists the right addresses, but ifconfig doesn't 1150647820 M * vrwttnmtu I don't know about IP 1150647824 M * vrwttnmtu it's not installed yet 1150647826 M * bonbons does ifconfig just parse files in /proc? 1150647828 M * vrwttnmtu Let me grab it 1150647842 M * daniel_hozac ifconfig uses netlink here to get IPv6 addresses. 1150647897 M * Bertl okay, off for dinner now, back shortly ... 1150647898 M * vrwttnmtu ip -6 addr shows only the host IP in the host 1150647903 N * Bertl Bertl_oO 1150647946 M * vrwttnmtu ip works, ifconfig doesn't. 1150647964 M * daniel_hozac very odd. 1150648000 M * vrwttnmtu I don't mind, particularly, but it's just something "that shouldn't be" 1150648001 M * vrwttnmtu :) 1150648028 M * vrwttnmtu daniel_hozac, I did run chbind6 with no args before 1150648033 M * vrwttnmtu I wonder if that's done something 1150648039 M * vrwttnmtu (Before you gave me the syntax) 1150648044 M * daniel_hozac it shouldn't. 1150648069 M * bonbons vrwttnmtu: need to look why, they probably take other ways inside of kernel to get information 1150648101 M * vrwttnmtu OK 1150648114 M * vrwttnmtu Well, I'm going to have a little play around with this v6 vserver 1150648120 M * vrwttnmtu Are there any known problems with it? 1150648212 M * daniel_hozac apparently ifconfig isn't isolated yet :) 1150648336 M * bonbons possibly more incomplete isolation issues, and undefined behavior (in relation to assigning packets to sockets) with IPv6 sockets that also cover IPv4 addresses (IPv4+IPv6 ANY, and IPv6 mapped IPv4) 1150648353 M * bonbons so make IPv6 socket IPv6only 1150648465 M * bonbons by the way, looks like IPv4 has some isolation issues tooin /proc/net/ 1150648888 M * bonbons I see host devices/addresses/net/{igmp,igmp6,ipv6_route,rt_cache} 1150649102 Q * MrX Ping timeout: 480 seconds 1150649299 J * MrX ~urk@219.95.2.229 1150649394 N * Bertl_oO Bertl 1150649395 M * trippeh Bertl_oO: Compiles fine noew 1150649397 M * Bertl back now 1150649397 M * trippeh now 1150649418 M * trippeh (thus it must be bug-free, right?) ;) 1150649457 M * trippeh Err, wait, I havn't gotten to that stage in the build yet. 1150649464 A * trippeh is overheating 1150649470 M * Bertl lol 1150649478 A * Wonka dumps some water on trippeh 1150649492 M * Bertl streamy :) 1150649500 M * bonbons Bertl: I see host devices/addresses/net/{igmp,igmp6,ipv6_route,rt_cache} 1150649531 M * bonbons do you know of head where to look for ifconfig's access to interface/address listing? 1150649550 M * Bertl netlink? 1150649562 M * Bertl or plain old ioctl? 1150649609 M * bonbons daniel said it is using netlink to query it's information... 1150649631 M * daniel_hozac that was just what the strace said. 1150649645 M * bonbons what's needed is place where isolation is insufficient 1150649651 M * daniel_hozac i haven't checked the source yet. 1150649743 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia' 1150649819 M * Bertl bonbons: have you compiled the debug stuff in? 1150649827 M * bonbons yep 1150649835 M * Bertl logging the addr checks might give you some clues, no? 1150649848 M * bonbons maybe 1150650052 M * bonbons got nothing from debug_net and debug_nid 1150650217 M * vrwttnmtu I think, speaking from a very uninformed view, that if that IPv6 patch (and a little interface interfaces/0/ipv6 was added), it'd be fine for testing. Perhaps an --enable-ipv6 and a menuconfig option for those that wanted it 1150650250 M * vrwttnmtu vserver + ipv6 = my number one wish 1150650252 M * vrwttnmtu :) 1150650254 M * Bertl bonbons: what flags did you set? 1150650304 M * vrwttnmtu Hollow, perhaps if the ipv6 USE flag is set, you could pull in the patch, etc? 1150650324 M * trippeh Bertl: http://paste.linux-vserver.org/103 1150650328 M * trippeh It got further ;) 1150650365 M * bonbons Bertl: 255 > debug_nid, 127 > debug_net 1150650529 M * Bertl trippeh: ah, yeah, that was fixed in mainline too .. sec 1150650558 M * waldi Bertl: hmm, what is rc23.1? 1150650588 M * Bertl a test release for rc24 (to be release pretty soon) 1150650598 M * waldi define "soon"? 1150650603 M * Bertl today 1150650606 M * waldi okay 1150650654 M * waldi 2.6.17 release for debian unstable is scheduled for tomorrow 1150650666 M * Bertl we should be ready for that 1150650752 M * trippeh waldi: Is the debian unstable kernel on 2.0.x or 2.1.x vserver patches? 1150650794 M * waldi 2.0 1150650811 M * trippeh Ok 1150651426 Q * m4z Quit: Lost terminal 1150651437 J * m4z m4z@bastard-operator.from-hell.net 1150653397 M * Bertl it seems marlow had an accident with the ml host ... 1150653426 M * Bertl (at least I have no better explanation for the amount of spam atm) 1150653539 M * gdm ok, was beginning to wonder if it was just me 1150653553 M * Bertl already sent him email .. 1150653605 M * bonbons Bertl: is task_get_nx_info(current) available when operating on a seq_file? 1150653628 M * Bertl yes, but it might not be correct 1150653640 M * Bertl especially when communicating via netlink 1150653678 M * bonbons is there a more reliable way to find out which context is requesting the data? 1150653703 M * Bertl the socket should be tagged 1150653726 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.17-vs2.0.2-rc23.2.diff 1150653731 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.17-vs2.1.1-rc23.2.diff 1150653743 M * Bertl trippeh: fixes the loop issue too 1150653758 M * Bertl the 2.1.1 test version has almost complete ocfs2 support :) 1150653847 M * bonbons Bertl: whould be for static int if6_seq_show(struct seq_file *seq, void *v) in net/ipv6/addrconf.c (around line 2670) 1150654238 J * m4z_ m4z@bastard-operator.from-hell.net 1150654242 Q * m4z_ Quit: 1150654483 Q * FireEgl Quit: Bye... 1150654538 J * aphetadus ~joaopaulo@201-26-50-74.dsl.telesp.net.br 1150654547 M * Bertl welcome aphetadus! 1150654594 M * aphetadus thanks Bertl! 1150654656 M * Bertl bonbons: from where is that used? 1150654755 M * Bertl bonbons: looks like file access, probably proc? 1150654776 M * Bertl (in this case you can assume that current is correct) 1150654792 M * bonbons yep, looks like proc 1150654911 M * bonbons hard to tell from where it's called as it's only called by address as set in a struct 1150655005 M * aphetadus Hey guys, I'm trying to understand how processes inside a guest are scheduled... Are all the processes in the machine are scheduled like a canonical linux machine (meaning all processes in a same schedule queue), or the processes are scheduled in a per-context order, or something like that? 1150655166 M * Bertl basically they are scheduled like on a normal linux system 1150655199 M * Bertl in addition, you can 'tune' the scheduler to do fair or hard limit scheduling 1150655224 M * Bertl i.e. limit the relative or absolute amount of cpu time spent for a guest 1150655230 M * aphetadus ok, this is done through vsched, right? 1150655251 M * Bertl yes, basically 1150655372 J * s0undt3c1 ~s0undt3ch@bl7-244-180.dsl.telepac.pt 1150655812 Q * s0undt3ch Ping timeout: 480 seconds 1150655965 Q * lilalinux Remote host closed the connection 1150656018 J * Planeti ~ircop@80.80.168.248 1150656027 M * aphetadus ok, thanks Bertl 1150656211 N * s0undt3c1 s0undt3ch 1150656348 J * AfterDeath ~AfterDeat@ppp-70-251-240-152.dsl.rcsntx.swbell.net 1150656384 P * AfterDeath Infected by the partial reasoning virus. Set your part message so that u may be infected!!! MUAHAHAHHA 1150657060 M * Bertl welcome Planeti! 1150657067 M * Bertl aphetadus: you're welcome! 1150657082 Q * teukka Quit: Terminated with extreme prejudice - dircproxy 1.0.5 1150657095 M * Planeti sorry 1150657111 M * Planeti speakk iltliano nesuno cui 1150657112 M * Planeti a no 1150657127 M * Bertl well, we speak english here :) 1150657150 M * Planeti ciao adeso 1150657152 M * Planeti bay 1150657166 Q * Planeti Quit: 1150657180 M * trippeh Hehe 1150657299 Q * Viper0482 Remote host closed the connection 1150657377 Q * yarihm Quit: Leaving 1150657497 J * matti matti@212.244.232.46 1150657517 M * Bertl welcome matti! :] 1150657561 M * matti Hello Bertl :-) 1150657712 P * aphetadus 1150658050 J * FireEgl Atlantica@Atlantica.US 1150659365 J * doener ~doener@i5387C029.versanet.de 1150659414 M * Bertl wb doener! 1150659424 M * doener hi 1150659757 M * bonbons vrwttnmtu: now I've a version where ifconfig does not leak information (still "old" kernel) 1150660114 M * Nam hey Bertl 1150660124 M * Bertl evening Nam! 1150660134 M * Bertl s/evening/morning?/ 1150660135 M * Nam you get my msg from last night? 1150660155 M * Bertl probably missed that, let's hear! 1150660161 M * Nam i fixed it 1150660168 M * Bertl AHA!? 1150660178 M * Nam it must have been the compiler 1150660194 M * Bertl ah, so we have an issue with 3.4.x! 1150660197 M * Nam it could have been something to do with glibc as well... but i can't be sure of that 1150660204 M * Nam yup, it would seem so 1150660228 M * Bertl can we double check that and keep a 3.3.x binary (of the tools) and make a 3.4.x one too? 1150660252 M * Nam as soon as I downgraded the distro back to 10.2, with the 3.3.6 gcc, it started working 1150660254 M * Bertl I would love to compare the strace or gdb output 1150660307 M * Nam i don't have a copy of them anymore 1150660447 M * vrwttnmtu bonbons, A version of the patch? 1150660470 M * bonbons yep, just uploaded 1150660487 M * vrwttnmtu OK - can you send me the link? 1150660505 M * bonbons http://homepage.internet.lu/brunop/vserver/ 1150660522 M * vrwttnmtu Where is .lu btw? 1150660595 M * bonbons in luxembourg, between france, germany an belgium 1150660627 M * vrwttnmtu Aaah, I know where Luxembourg is, but I was thinking around Latvia, Lithuania, etc. 1150660709 M * vrwttnmtu bonbons, Supporting France, or Germany in the cup? Or non at all? 1150660734 M * bonbons I have #undef cup ;) 1150660737 M * vrwttnmtu You have a french name, so I am guessing... 1150660747 M * vrwttnmtu :) 1150660772 M * waldi Bertl: hmm, did you add the change for the bindmounts? 1150660829 M * Bertl waldi: which one? 1150660836 M * waldi nodev 1150660846 M * Bertl no, url? 1150660878 M * waldi hrm 1150660918 M * bonbons vrwttnmtu: football does not interest me at all, so except having heard that there is the worldcup I know nearly nothing about it 1150661013 M * vrwttnmtu bonbons, Can't avoid it in England. 1150661032 M * Bertl bonbons: ah, you heard about the worldcup too, so it must be happening out there, somewhere :) 1150661066 M * vrwttnmtu Bertl and bonbons both come from countries with no World Cup team... :) 1150661121 M * Bertl and we are sooo sorry about that :*G* 1150661136 M * bonbons vrwttnmtu: even if there was a team the state would be exact the same! stdin | grep -v football | processing 1150661180 M * vrwttnmtu I can understand people not supporting local teams playing sports - but if it's your nation - you have to get behind them :) 1150661203 M * vrwttnmtu I expect you're both too young :) 1150661241 M * Bertl probably, will check again with sixty :) 1150661261 M * daniel_hozac come on. football has to be one of the most boring sports ever. 1150661279 M * vrwttnmtu If you just look at it as 22 men kicking a white thing around, yes 1150661282 M * vrwttnmtu Same as cricket 1150661301 M * vrwttnmtu But if you get into it, the emotion, the psychology of it, it's great 1150661310 M * Bertl we have one ... outside in the garden ... 1150661310 M * vrwttnmtu I used to hate football 1150661346 M * vrwttnmtu daniel_hozac, What's a good sport then? 1150661350 M * bonbons daniel_hozac: that's the reason why it has such success ;) 1150661361 M * Bertl vrwttnmtu: billards e.g. snooker 1150661362 M * daniel_hozac one where things happen more often than every 5 minutes, at least :) 1150661379 M * vrwttnmtu Bertl, I like snooker, for sure 1150661400 M * vrwttnmtu daniel_hozac, It's the build-up, the pressure, the excitement, the disappointment, not just the goals 1150661443 M * Bertl they use pressure? ... ah yes, for the ball :) 1150661493 M * vrwttnmtu The problem is, daniel_hozac, that if something happens < every 5 minutes, it becomes worthless. Like in basketball - it's just end to end - scoring every time. I prefer cricket - it can be a whole morning, standing in the sun, getting hot, and annoyed, and then suddenly, the game can turn 1150661544 M * daniel_hozac i guess i just have a short attention span. 1150661570 M * daniel_hozac now, i don't mind playing football (except for all the running, heh), just the watching. 1150661606 M * Bertl short what? what are we talking about? 1150661611 M * vrwttnmtu So you don't mind if we beat you 5-0 on Tuesday, daniel_hozac ? 1150661627 M * vrwttnmtu (Won't happen, but...) 1150661629 M * daniel_hozac absolutely not. in fact, that would make me happy :) 1150661640 M * vrwttnmtu Are you anti-Swedish then? :) 1150661652 M * daniel_hozac when it comes to sports, yes. :) 1150661657 M * vrwttnmtu Why? :) 1150661662 M * vrwttnmtu I'm curious 1150661695 M * daniel_hozac because it keeps people from getting all excited about winning. 1150661714 M * vrwttnmtu Well, let's ban all art then. And music. 1150661715 M * vrwttnmtu :) 1150661723 M * vrwttnmtu Equilibrium, here we come! :) 1150661745 M * daniel_hozac i just think it's silly to get excited about other people playing sports. 1150661762 M * vrwttnmtu They're other people, sure, but they're **your** other people. 1150661766 M * vrwttnmtu Do you watch films? 1150661774 M * daniel_hozac all the time. 1150661777 M * vrwttnmtu It's silly to get excited about other people on a film. 1150661797 M * daniel_hozac not quite the same thing. films are about escapism. 1150661803 M * daniel_hozac surely that is not the case for sports? 1150661829 M * vrwttnmtu I don't know - perhaps it is - 90 minutes of cheering your team on 1150661874 M * vrwttnmtu cp arch/x86_64/boot/bzImage /boot/2.6.16.20-vs2.1.1-rc22IPv6 1150661877 M * vrwttnmtu Time for a rebot 1150661878 M * vrwttnmtu reboot 1150661879 M * vrwttnmtu :) 1150661885 M * vrwttnmtu You'll all be sorry to hear 1150661886 M * vrwttnmtu :) 1150661893 Q * vrwttnmtu Quit: Leaving 1150662128 J * vrwttnmtu ~eryktyktu@82-69-161-137.dsl.in-addr.zen.co.uk 1150662139 M * vrwttnmtu 2.6.16.20-vs2.1.1-rc22IPv6 1150662174 Q * mnemoc Ping timeout: 480 seconds 1150662199 M * vrwttnmtu So bonbons, do I still have to add the IP to the host, before assigning it to the guest? 1150662224 M * bonbons the order of assigning has not importance at all 1150662228 M * vrwttnmtu OK 1150662249 M * vrwttnmtu So, just to understand it - the host answers all arp requests for the guest IPs, is that right? 1150662265 M * bonbons they guest can only see/use it when it's assigned to a device, but you can tell the kernel to allow use once the IP as configured on a device 1150662294 M * vrwttnmtu " you can tell the kernel to allow use once the IP as configured on a device" - explain? 1150662310 M * bonbons yes, the host does all the black-magic to get connections working (and for IPv6 there is no ARP, just ICMPv6) 1150662323 M * vrwttnmtu aah yes, of course :) Doh! 1150662366 M * vrwttnmtu Sweet. I only see my guest v4 and v6 address in the vserver now 1150662386 M * vrwttnmtu Is it possible to run an "only v6" guest? There's no requirement for v4? 1150662388 M * bonbons you assign guest ip 2000::1, but is't not set on and interface. As soon as an interface gests ip 2000::1 your guest can use it (e.g. some service had bound to ::) 1150662420 M * bonbons what you need just depends on userspace apps 1150662437 M * vrwttnmtu OK - there's nothing in the vserver internals that requires a v4 address? 1150662448 M * bonbons if something prevents you from running IPv6 in a guest without IPv4 then it must be fixed 1150662454 M * vrwttnmtu Is there a v6 wiki page? 1150662467 M * bonbons v6 wiki page? 1150662478 M * vrwttnmtu A wiki page about running a v6 vserver 1150662495 M * vrwttnmtu If not, I will write up my laymans view of getting it working 1150662500 J * mnemoc ~amery@kilo105.server4you.de 1150662502 M * bonbons a page in the wiki explaining how to use IPv6 for vserver, not that I know of 1150662544 M * Bertl nap attack ... back later ... 1150662553 N * Bertl Bertl_zZ 1150662836 M * trippeh Sigh, I need to get the dual dual core opteron build machine on-line, compiling 18 kernel images on a amd64 3000+ is taking the entire evening 1150662878 M * bonbons trippeh: 18? what do you do with all those kernel images? 1150662934 M * daniel_hozac wow, 18 is even worse than me. 1150662964 M * trippeh bonbons: images for Sarge, Debian Unstable + i386, amd64, then there's the variants, smp, up, and their 100hz (for vmware, sight) counterparts 1150662997 M * trippeh s/sight/sigh 1150663038 M * trippeh Just cut it down a bit, due to i386 SMP alternatives (live patching of UP vs SMP code in the kernel) 1150663069 M * trippeh Hopefully amd64 SMP alternatives will be accepted soon, that would rock ;) 1150663073 M * vrwttnmtu bonbons, http://linux-vserver.org/IPv6 1150663161 M * trippeh For 2.6.16 I had 24 images in total, so 18 is a major step ahead 1150663212 M * bonbons trippeh: that's still a lot! Don't you have a compile farm available for that kind of job? 1150663270 M * bonbons vrwttnmtu: good start :) 1150663307 M * vrwttnmtu bonbons, If I ask the questions, I'm sure someone else will. 1150663308 M * vrwttnmtu :) 1150663324 M * vrwttnmtu I must say though, I do find it hard to find information about vserver on the wiki 1150663330 M * trippeh bonbons: x86_64 smp alternatives will cut it down to 12, 3 images for each arch/dist combination 1150663344 M * vrwttnmtu And I couldn't find the experiment patches today either. 1150663399 M * trippeh bonbons: No.. I find it easier to do this on my private systems rather than going through the massive byrocracy to get something up at work 1150663447 M * bonbons trippeh: that's understandable! Burocracy is always a pain 1150663490 Q * Nam Remote host closed the connection 1150663530 M * bonbons vrwttnmtu: for where to find information you need to ask Bertl or Daniel, checking who is quicker ;) 1150663554 M * vrwttnmtu Yeah, that's what I do 1150663554 M * vrwttnmtu :) 1150663597 M * trippeh The only way outside the burocracy is to classify it as a top secret skunk works project... ;-) 1150663628 M * bonbons hehe 1150663671 M * bonbons I would just schedule it for night-work (menuconfig everything as needed an then issue build right before leaving) 1150663721 M * daniel_hozac that's what i used to do back in the day of 18 hour builds. 1150663846 M * bonbons and the rest is done in background with some niceness to not distrub too much 1150663930 M * trippeh Our security guidelines forbids building on servers doing actual work, so niceness is not a problem ;) 1150663996 M * bonbons don't you have workstations or non-production servers? 1150664022 M * vrwttnmtu trippeh, What's the reasoning for that? 1150664078 M * trippeh vrwttnmtu: We've had attacks where builds have been manipulated 1150664097 M * vrwttnmtu Wow. By who? 1150664208 M * vrwttnmtu Package maintainers, people in your company? 1150664246 M * trippeh I'm not classified to know ;-) 1150664257 J * Nam ~nam@70.78.64.62 1150664527 M * trippeh I'm walking on the edge by doing a few builds on private systems, it's "fine" for now as its not (yet) for production use 1150664729 M * vrwttnmtu Does vhashify "just work" once it's enabled? I mean, it finds the common files, and sorts everything out itself? Or does the administrator have to tell the system which files are the same as other files? 1150664973 M * trippeh vrwttnmtu: I wish I could tell more about that particular incident, the cleanup op was *really* interesting. 1150664992 M * vrwttnmtu trippeh, Bombs dropped in foreign countries? :) 1150664994 M * trippeh Spent like 30-40 million USD in one week 1150665024 M * vrwttnmtu Not that guy from the UK who was looking for UFO evidence? 1150665068 Q * shedi Quit: Leaving 1150665159 M * trippeh Hehe, no 1150665194 M * trippeh This was in the private sector 1150665204 M * vrwttnmtu That's a hell of a bill for a cleanup 1150665210 M * vrwttnmtu I'd do it for half that. :) 1150665537 M * trippeh Sure you could ;) 1150665608 M * vrwttnmtu I'll even split it with you 1150665630 M * derjohn trippeh, i offer 60/40 1150665638 M * vrwttnmtu Damn you, derjohn ! 1150665638 M * derjohn :) 1150665645 M * vrwttnmtu OK, let's just do 33/33/33 1150665664 M * vrwttnmtu derjohn, Are you the hints_from_john guy? 1150665665 M * derjohn thats 333+333=666! OMG ! 1150665668 M * trippeh I got my bonus ;-P 1150665684 M * derjohn yes, am i the man who writes the hints :) 1150665692 M * vrwttnmtu Excellent :) 1150665702 M * vrwttnmtu Can you tell me about vhashify? Does vhashify "just work" once it's enabled? I mean, it finds the common files, and sorts everything out itself? Or does the administrator have to tell the system which files are the same as other files? 1150665707 A * derjohn is just so lazy placing the stuff in FAQ .... 1150665735 A * derjohn wants to make is name known to the vserver world ;) 1150665770 M * derjohn vrwttnmtu, I cannt tell you in detail about vunify or vahsify, because I didnt use it yet. just in theory. 1150665781 M * vrwttnmtu Aaah. 1150665812 Q * bonbons Quit: Leaving 1150665847 M * derjohn vrwttnmtu, you have to build the hashes once to "hashify" all your guests. the checksums are saved somewhre (.vhashify or so). 1150665885 M * vrwttnmtu OK - sounds OK 1150665893 M * vrwttnmtu So you just run it from time to time 1150665898 M * vrwttnmtu Whenever files might be updated 1150665907 M * vrwttnmtu Cron @ 5.30 am... 1150665913 M * derjohn nope! it has to be done once. 1150665924 M * vrwttnmtu But when the files change? 1150665929 M * derjohn after that you have a gigantic "hardlink-farm" 1150665941 M * derjohn same "files" point to the same inode. 1150665945 M * trippeh vrwttnmtu: The cleanups do become expensive when one gets owned with *no* security whatsoever. The people in charge was of the old mainframe crowd, where telnet and client side security reigns 1150665950 M * trippeh (they are fired now.) 1150665994 M * derjohn if a guest overwrites the file, the inode gets copiesd (new: CoW) and this guest gets a own inode for the file. 1150665997 M * vrwttnmtu trippeh, I wince when I see people logging into boxes, routers, and switches with telnet. "It's OK, it's a safe network", they say. Fools. 1150666004 Q * brc_ Ping timeout: 480 seconds 1150666023 M * vrwttnmtu derjohn, Sounds like just an extra level of things that can go wrong :) 1150666034 M * vrwttnmtu I'll just buy more disks. :) 1150666061 M * derjohn vrwttnmtu, well it has several advantages. not only the space. every thought about inode caching (buffers)? 1150666070 M * trippeh vrwttnmtu: Thats what they said too. 1150666078 M * trippeh They're not saying it anymore, weird ;-) 1150666078 M * vrwttnmtu derjohn, Nope.. :) 1150666108 M * derjohn well, in a hashified system you have to use only one cache for all guests who "share" the inode. 1150666114 M * vrwttnmtu trippeh, Yeah, it makes me mad. I almost think it would be better if all networks ran with public IPs, and no firewalls. That would force the use of sane security on the hosts. 1150666116 M * derjohn (the unified file) 1150666158 M * derjohn vrwttnmtu, backups will be drastically smaller, to you dont put much load on the boxen when rsyncing the host 1150666181 J * Aiken ~james@tooax6-025.dialup.optusnet.com.au 1150666212 M * trippeh vrwttnmtu: Funny how the firewalls was of no use, with thousands of workstations on the inside. 1150666286 M * derjohn think about 30 guest with 3 GB .. that's 90 GB for the backup and assumed the OS has 2 GB unified files: 32 GB ... 1150666302 M * daniel_hozac vrwttnmtu: hashify will do everything for you. assuming you don't have broken package management inside the guest, you shouldn't have any problems. 1150666342 M * vrwttnmtu daniel_hozac, Question: What do people running vservers for 3rd parties do - rely on the clients to keep their stuff updated, or update them for them? 1150666356 M * derjohn daniel_hozac, your discussion with micah .. was that a broken package? 1150666369 M * trippeh vrwttnmtu: I started with the company 6 months before it exploded, telling relevant people how this would blow up in their faces sooner than later 1150666384 A * trippeh loved beeing so right ;) 1150666422 M * vrwttnmtu trippeh, It was you :) 1150666430 M * trippeh Ha 1150666472 M * vrwttnmtu Proving them wrong. :) 1150666540 M * Nam which define line specifies how many ips can be assigned per vserver? 1150666626 M * doener su - 1150666631 M * doener uhm, oops :) 1150666635 M * vrwttnmtu Heh :) 1150666638 M * vrwttnmtu Password: 1150666645 M * doener secret 1150666647 M * trippeh :-D 1150666659 M * doener oh, guess I have to change that now ;) 1150666673 M * vrwttnmtu Just make it s3cr3t 1150666683 M * vrwttnmtu Or password1 1150666692 M * vrwttnmtu No-one ever bothers to try those any more 1150666813 M * trippeh masterkey 1150666844 M * Nam NB_IPV4ROOT 1150667170 M * doener ah, zeng finally published his MoreUbuntu HowTo :) 1150667501 A * mugwump caught a sysadmin using "passw0rd" once 1150667525 M * mugwump hey, quick kernel question for anyone around who might know, 1150667539 M * mugwump ever heard of "Resource Groups" ? 1150667609 M * doener yep, but doesn't trigger anything right now... got any more details? 1150667650 M * mugwump http://lkml.org/lkml/2006/6/18/21 1150667681 M * mugwump I'm wondering if this is some standard feature, like process groups 1150667754 M * doener http://lwn.net/Articles/181857/ 1150667770 M * mugwump ahhh 1150667778 M * daniel_hozac derjohn: hmm? when? 1150667884 M * derjohn daniel_hozac, some time ago, micah ask about vhashify a lot .. I guess it working now in Debian (not that I encounter a pitfall when trying vhashify :)) 1150667932 M * derjohn argh: 2.6.17 says: ACPI: unable to load system description tables ..and doesnt load the kernel :( anyone else had that problem, too? 1150667945 M * daniel_hozac ah yes, the strange "more disk usage after hashification" thing. 1150667969 M * daniel_hozac i don't think we ever managed to track that one down. 1150667985 M * derjohn daniel_hozac, oh ... 1150668004 M * derjohn daniel_hozac, well, currently i have other things to do :( *grr* 1150668081 M * daniel_hozac well, as long as you're not seeing it, i guess it does not matter much :) 1150668093 A * vrwttnmtu isn't touching vhashify now :) 1150668255 M * doener daniel_hozac: I assume that it happens when the hashification is done while the vserver is running 1150668276 M * doener because the files in /vservers/.hash are _new_ files and the files inside the vserver are replaced 1150668278 M * daniel_hozac doener: yeah. 1150668296 M * doener ie. as long as they are in use, the disk usage is higher because you have the old and the new file 1150668354 M * daniel_hozac that seems like such an obvious explaination though. i wonder why that wasn't the conclusion back then, or was it? 1150668363 M * daniel_hozac derjohn: got a timestamp? 1150668371 M * doener and besides that, it breaks debian's start-stop-daemon ;) 1150668411 M * mugwump unify-dirs never increased disk usage :-P 1150668509 M * daniel_hozac doener: hmm? which debian? 1150668535 M * doener sarge IIRC... 1150668572 M * doener after "vserver foo hashify", while "foo" was running, start-stop-daemon refused to stop my lighttpd, because the file wasn't the same anymore 1150668572 M * Nam are there any tools to say exactly what cpu usage and memory usage a particular vserver is using? 1150668609 M * derjohn daniel_hozac, maybe: http://irc.13thfloor.at/LOG/2006-04/LOG_2006-04-20.txt 1150668718 M * daniel_hozac Nam: cat /proc/virtual//{limit,sched}? 1150668735 M * daniel_hozac doener: ah right, that. i guess we can't really fix that one. 1150668807 M * doener daniel_hozac: I'll probably check the start-stop-daemon source next week and file a bug/write a patch... It should be able to deal with the case that the file has been deleted 1150668811 M * Nam daniel_hozac: arn't those settings, not stats? 1150668830 M * daniel_hozac Nam: they're both. 1150668833 M * Nam like when you do "uptime" and get the load average, every vserver will report the same thing 1150668840 M * daniel_hozac virt_uptime. 1150668850 M * daniel_hozac or /proc/virtual//cvirt 1150668895 Q * f_ Quit: This computer has gone to sleep 1150668924 M * Nam NICE, exactly what I was looking for, thx 1150668931 M * daniel_hozac doener: hmm, du wouldn't account for old files, would it? 1150668977 M * Nam is there any documentation on what each entry/column is for? 1150669036 M * daniel_hozac http://linux-vserver.org/HowTo+Read+ProcFS 1150669128 M * doener daniel_hozac: true 1150669151 M * Nam man, your guys rock, thank again daniel_hozac 1150669308 M * vrwttnmtu Good night all 1150669313 M * daniel_hozac good night. 1150669320 Q * vrwttnmtu Quit: Leaving 1150669917 M * Nam is there any reason why adding a "context" file and assigning an id, i get the following error... "chbind: vc_net_create(): Invalid argument", but when i remove the context file, it works 1150670030 M * Nam even if i start a vserver, view the context id it was automatically assigned, then putting that in the context file, still, same error 1150670050 M * Nam hmm... 1150670052 M * daniel_hozac you can't assign dynamic ids. 1150670054 M * daniel_hozac they're dynamic. 1150670055 M * Nam i set it to 2, and it worked 1150670057 M * daniel_hozac 2-49151. 1150670061 M * Nam ah 1150670067 M * Nam thx 1150670099 M * Nam 49152-65536 are automatic ids or something? 1150670166 J * id23 ~id@p50813C68.dip0.t-ipconnect.de 1150670179 M * id23 Hallo #vserver 1150671232 J * dna ~naucki@dialer-180-70.kielnet.net 1150671359 Q * dna Quit: 1150672744 J * darkordek1 ~darkordek@85.100.185.14 1150673342 Q * id23 Quit: Leaving 1150673851 P * darkordek1 1150673894 Q * derjohn2 Ping timeout: 480 seconds 1150673910 J * derjohn2 ~aj@dslb-084-058-208-147.pools.arcor-ip.net 1150674509 J * shedi ~siggi@130.208.221.254 1150675193 Q * derjohn2 Quit: Verlassend