1206749619 J * DanielW ~danielw@f050064218.adsl.alicedsl.de 1206749620 M * DanielW hi 1206749642 M * DanielW must a guest set a default route? 1206749643 Q * xdr Ping timeout: 480 seconds 1206749655 M * DanielW or does the host system does that for the guest? 1206750409 Q * Piet Quit: Piet 1206751298 N * Bertl_zZ Bertl 1206751321 M * Bertl DanielW: guest do not set routing, the host does that 1206751327 M * Bertl *guests 1206751361 M * DanielW mhh ok, than i have no idea why my vserver is not responding after rebooting 1206751419 M * Bertl the guest or the host? 1206751421 M * DanielW i am trying to do that: untar an nearly complete / filesystem with archlinux in it (with new /etc/init.d/rc) to a remote linux vserver based vserver 1206751443 M * DanielW untar works and i can run pacman then on the vserver (pacman is the archlinux package manager) 1206751456 M * Bertl sounds good 1206751463 M * DanielW but after reboot it doesn't come up 1206751483 M * Bertl means, the guest startup fails in some way, yes? 1206751492 M * DanielW or well i don't know. it doesn't answer to icmp echo requests 1206751506 M * DanielW yeah i think that, but don't know. 1206751519 M * Bertl so, from the way you are describing that, I assume you have no control of the host, yes? 1206751526 M * DanielW correct 1206751551 M * Bertl so, a very important detail is how the guest is started, which usually depends on the distribution 1206751570 M * Bertl what was the 'original' distro on that guest? 1206751575 M * DanielW debian 4.0 1206751607 M * Bertl okay, can you somehow remote re-install that? 1206751611 M * DanielW sshd and crond and init and bash the only prozesses in ps output 1206751613 M * DanielW i can 1206751620 M * DanielW have debian back on it 1206751630 M * DanielW i am just reuploading my archlinux tar archive 1206751648 M * DanielW but i don't won't to reboot it again without any changes 1206751656 M * Bertl okay, while still on debian, check if you have a real init or a blend through init 1206751666 M * DanielW who can i do that? 1206751668 M * DanielW how 1206751685 M * Bertl the easiest way would be to tell init to change the runlevel 1206751704 M * Bertl if init compies and changes the runlevel, you have a real init running 1206751708 M * Bertl *complies 1206751768 M * DanielW init: /dev/initctl: No such file or directory 1206751776 M * DanielW so what does that mean? 1206751795 M * Bertl thought so, that means the guest is using a sysv init style 1206751810 M * Bertl which does not start an actual init for the guest 1206751831 M * DanielW so the host just execudes /etc/init.d/rc to init the system? 1206751860 M * Bertl yes, but that should not be a big problem for you, you just have to make sure that this works as expected, and leaves at least one process running 1206751898 M * Bertl I'd suggest to install Linux-VServer somewhere on a machine under your control (can be a virtual one too :) 1206751901 M * DanielW http://linux-vserver.org/Howto_make_bsd-style_init_sysv-compatible <- i tried that 1206751954 M * DanielW yeah i thought about that too. what would be the fastest way to set up a linux in qemu-kvm with vserver installed? 1206751983 M * Bertl IIRC, there is a knoppix CD with Linux-VServer support 1206751983 M * DanielW there isn't a thing like live linux-vserver cd? 1206751992 M * DanielW oh 1206752127 M * DanielW you have no idea where to find that? 1206752183 M * Bertl I have no idea which version included Linux-VServer support, maybe the current version still does 1206752219 M * Bertl google says, Knoppix 5.2 should be Linux-VServer enabled :) 1206752278 M * Bertl and I have a qemu line for you :) 1206752294 M * Bertl qemu -boot d -cdrom /path/to/KNOPPIX.iso -user-net -enable-sound 1206752609 M * DanielW there is no 5.2? 1206752691 M * Bertl 5.2.0 maybe? 1206752698 M * DanielW ok the 5.3 dvd has it. but it is 4.5 GB 1206752740 M * Bertl well, it's probably easier to boot a Linux-VServer enabled kernel on your real hardware then :) 1206752846 M * Bertl I can offer you to upload your guest somewhere (where I can reach it) and I'll give it a spin, and tell you what part of the startup fails 1206752880 M * DanielW ok found a physical DVD here it is labeled "knoppix 5.2" :-) 1206752891 M * Bertl even better :) 1206752992 Q * ftx__ Ping timeout: 480 seconds 1206753358 M * DanielW it doesn't boot :-( i enter vserver to select vserver kernel and then black screen :-( 1206753414 M * DanielW ok without kvm it works 1206753434 M * Bertl 64bit machine? 1206753445 M * DanielW 32 bit 1206753479 M * DanielW core 2 duo t7500 but using 32 bit os 1206753638 M * DanielW ok now i should learn how to setup a vserver 1206753656 M * Bertl that is quite simple, and for your (debian) guest you can do that via network 1206753667 M * Bertl first check that the networking works as expected 1206753729 M * DanielW i can browse websites in konqueror 1206753745 M * Bertl good, now do something like this (adjust the ugly details :) 1206753748 M * Bertl vserver etch32 build -m debootstrap --context 10104 --hostname etch32.debian.org --interface eth1:10.1.4.32/24 -- -d etch -m http://ftp.debian.org/debian -- --arch i386 1206753768 M * DanielW ok 1206753790 M * Bertl if the installation is a little outdated, you might first want to update util-vserver from backports 1206753874 M * DanielW 0.30.211 is that outdated? 1206753891 M * Bertl pretty much, we are at 0.30.215 atm, 0.30.214 would be fine 1206754813 M * DanielW oh apt-get install is sooo slow when reading db from dvd 1206755144 M * Bertl probably not as slow as yum :) 1206755489 Q * hparker Remote host closed the connection 1206755664 M * DanielW ok don't get the newer util-vserver installed 1206755684 M * Bertl did you add the backports repository= 1206755687 M * Bertl s/=/? 1206755695 J * chealer_ ~chealer@dsl-67-55-13-20.acanac.net 1206755700 M * DanielW yes 1206755708 P * chealer_ 1206755712 M * Bertl and, what's the problem? 1206755793 M * DanielW well 1206755826 M * DanielW it needs a never libc6 1206755841 M * DanielW newer 1206755857 M * Bertl that's bad, especially as the tools do not really use glibc 1206755897 M * DanielW is there a only that packages no deps option in apt? 1206755908 M * Bertl well, did you try with the installed 0.30.211 tools? 1206755918 M * DanielW no will try that now 1206755958 M * DanielW how can i select where it is installing the filesystem for the virtual server? 1206755983 M * Bertl vserver - build --help 1206756640 M * DanielW ok it seems to download a debian system 1206756714 M * Bertl good, if that finishes, try to start/stop it 1206756828 M * DanielW i think that ftp.debian.org is slow 1206756843 M * Bertl that could as well be ... 1206756876 M * DanielW around 100 kilobyte/s 1206757170 M * Supaplex you know debian.org has a few *cough* mirrors 1206757200 M * DanielW ok now i can not start the vserver. it can not read /proc/uptime 1206757317 M * Bertl you need to run vprocunhide 1206757332 M * Bertl it seems the knoppix startup forgot that 1206757358 M * DanielW no vprocunhide here 1206757377 M * Bertl it's not in the normal path 1206757430 Q * julius_ Remote host closed the connection 1206757609 M * DanielW ok found it. but no it complains about missing /bin/umount. but /bin/umount it there 1206757636 M * Bertl who complains and when you do what exactly? 1206757651 M * DanielW vserver etch32 start 1206757680 M * Bertl so that is probably caused by the ancient tools then 1206757684 M * DanielW and then /usr/lib/util-vserver/vserver.functions line 1223 1206757731 M * Bertl try getting at least 0.30.212 (which is the debian stable!! default) 1206757734 M * DanielW and he is saying that he can not find /etc/vservers/etch32/vdir but that is there to 1206758195 M * DanielW i don't undersand debians apt 1206758224 M * Bertl me neither ... but usually there are debian folks around who do :) 1206758238 M * DanielW it don't update util-verser at all. it always needs an libc which it can not get 1206758266 M * Bertl you could try to build a recent version yourself? 1206758307 M * DanielW on that dvd booted virtual system. 1206758321 M * DanielW ok it may work 1206758591 M * DanielW internal configure error :-/ 1206758606 Q * yarihm Quit: Leaving 1206758645 M * Bertl no idea on that one 1206758707 M * Bertl maybe force the 0.30.214 sarge backports package in with dpkg? 1206758857 M * DanielW yeah will try that 1206758941 M * Bertl I will make a mental note that knoppix 5.2 seems broken 1206759088 M * DanielW i think it is not designed to allow bigger updates. but the existing version of utils-vserver should work 1206759110 M * Bertl yes, but as you demonstrated, it doesn't :) 1206759207 M * DanielW maybe its just me ;-) 1206759832 M * DanielW now with 0.30.214 the same error just a few lines later in that file 1206759840 M * DanielW it claims not to find /bin/umount 1206759867 M * Bertl and /bin/umount is there? maybe it'S a symlink or so? 1206759911 M * DanielW file /bin/umount ELF 32-bit LSB executable 1206759935 M * Bertl and ls -la /bin/umount says? 1206759979 M * DanielW normal setuid file 1206760009 M * Bertl okay, run the start with --debug and upload the entire output somewhere 1206760015 M * DanielW /usr/sbin/vserver-info it can not find that too. but it is there 1206760032 M * Bertl okay, strace -fF the /usr/sbin/vserver-info then 1206760046 M * Bertl (that's probably easier to track down) 1206760100 M * DanielW eh vserver-info seems to work without any prboelms 1206760114 M * DanielW but the vserver etch32 start thing says it can not find that 1206760186 M * Bertl could be all kind of issues, starting with a broken environment (path) over improper permissions on / to badly misconfigured tools 1206760206 M * DanielW it does not try to umount /ramdisk does it?!? und /UNIONFS 1206760215 M * DanielW why does it try to do that? 1206760231 M * Bertl does it? if so, it should be simple to fix 1206760266 M * DanielW debug looks like that. but i am not sure 1206760272 M * DanielW will upload the debug output 1206760284 M * Bertl in /etc/vservers/.defaults put a file called 1206760289 M * Bertl namespace-cleanup-skip 1206760301 M * Bertl which contains a list of pathes not to be touched 1206760317 M * Bertl actually, touching a file called 1206760323 M * Bertl nonamespace-cleanup 1206760327 M * Bertl there should be enough 1206760411 M * DanielW with the -skip or without? 1206760421 M * Bertl without the skip 1206760437 M * DanielW doesn't change anything 1206760477 M * Bertl well, probably depends on the util-vserver version 1206760486 M * Bertl it definitely works for the latest tools 1206760549 M * DanielW well that all is weird it claims to not find files which are there for sure 1206760564 M * Bertl do the --debug run and upload the output 1206760632 Q * arachnist Remote host closed the connection 1206760644 J * arachnist arachnist@plzdie.kthxbai.pl 1206760732 M * DanielW http://rafb.net/p/vwfYHd75.html 1206760898 M * Bertl check what's in /usr/lib/util-vserver/defaults/environment 1206761010 M * Bertl please do 'touch /etc/vservers/etch32/nonamespace-cleanup' 1206761013 M * DanielW PATH with /usr/bin /usr/sbin /bin and /sbin 1206761083 M * Bertl (and try to start the guest :) 1206761109 M * DanielW yeah that worked 1206761120 M * DanielW what was he doing? 1206761132 M * Bertl unmounting parts of the knoppix mounts 1206761145 M * Bertl you probably just got the path for the defaults wrong in the first try 1206761200 M * Bertl (or the tools do not support the global for that) 1206761218 M * DanielW thank you for your help :-) 1206761230 M * Bertl you're welcome! 1206761238 M * Bertl so the guest now starts and stops, yes? 1206761247 M * DanielW yes 1206761263 M * Bertl okay, make a copy with the 'clone' build method 1206761281 M * Bertl make sure to assign ip and such in this process 1206761304 M * Bertl you can then unpack your 'new' guest there via ssh/rsync/whatever 1206761316 M * Bertl (i.e. do it the same way you do it on your remote host) 1206761367 M * DanielW don't unterstand that ip thing but i think it is working 1206761390 M * DanielW i mean that ssh into the guest 1206761394 M * Bertl basically you specify the same attributes to the clone build method you did to the debootstrap build method 1206761416 M * Bertl (except for the debootstrap specific options) 1206761446 M * DanielW oh i just though of try to make my etch32 guest to a archlinux install the same way as i am trying it on the remote machine? 1206761461 M * DanielW why doing a new build? 1206761484 M * Bertl vserver arch build -m clone --context 10105 --hostname arch.debian.org --interface eth1:10.1.4.33/24 1206761498 M * Bertl something like that, it will preserve your original etch guest 1206761530 M * Bertl (again, adjust the details) 1206761535 M * DanielW but why? 1206761549 M * DanielW ah it does make a copy of my etch32 ? 1206761552 M * Bertl so that you can try your 'migration procedure' more than once 1206761576 M * Bertl yes, ah I forgot, you want to add 1206761590 J * dowdle ~dowdle@67-42-224-198.blng.qwest.net 1206761596 M * Bertl -- -d etch --source etch32 1206761713 M * DanielW i don't have enough diskpace for that. i am not sure if there is a way to get more at runtime of the qemu virtual pc 1206761722 Q * bronson Read error: Connection reset by peer 1206761744 M * Bertl don't think so 1206761760 J * bronson ~bronson@adsl-68-122-117-135.dsl.pltn13.pacbell.net 1206761791 M * Bertl but you can make a tar over network as well (or use rsync) 1206761921 M * etteyafed DOes anyone know if daniel_hozac's 2.6.22.19-vs2.3.0.34.1 kernel supports xen? 1206761945 M * Bertl what do you mean by that? 1206761963 M * etteyafed I am an idiot and don't know how to find out if there is xen support built into that kernel 1206761991 M * Bertl dom0 or domU and what xen version/type? 1206762019 M * Bertl in general, I'd say it should work on a harware virtualization xen domU 1206762050 M * Bertl but I'd consider it unlikely that it is compiled for dom0 or paravirtualization 1206762084 M * etteyafed Oh, no I might need to test a XEN guest on the 2.6.22.19-vs2.3.0.34.1 kernel (as a host) 1206762130 M * etteyafed SO yeah. It might work 1206762144 M * Bertl then you definitely need to boot the hypervisor and you want a dom0 kernel 1206762267 M * etteyafed Yeah. Well would have assumed that it would have worked in domU, most things do. I may have other options before that anyway. There are worse things than building your own kernel though. 1206762291 M * Bertl I'd say it works in domU, if you don't use paravirtualization :) 1206762338 M * DanielW Bertl: no should i try to restart that guest after i untared the arch "image" in it? 1206762349 M * Bertl yes 1206762349 M * DanielW is there a way to see what happens while booting? 1206762361 M * etteyafed I think my issues with SELinux may stem from the userspace AVC in the vserver being a different version than what workes with the kernel AVC in the host. 1206762369 M * Bertl DanielW: check that the init style is sysv for the guest 1206762399 M * Bertl /etc/vservers/vserver-name/apps/init/style 1206762432 M * DanielW that file is not there 1206762446 M * Bertl then create it with 'sysv' as contents 1206762480 M * DanielW ok 1206762488 M * DanielW now shutdown and than restart? 1206762499 M * DanielW does it show me that output of booting? 1206762516 M * Bertl hmm, maybe remove that file again before shutting down :) 1206762569 M * Bertl etteyafed: AVC? 1206762605 M * DanielW ok shutdown in guest doesn't work because of missing /dev/initctl 1206762620 M * Bertl which is the same as in your real guest, yes? 1206762625 M * DanielW yeah 1206762640 M * etteyafed Bertl: Access Vector Cache 1206762650 M * DanielW both with the debian and the arch files 1206762667 M * Bertl okay, so now stop the guest from the host 1206762672 M * Bertl and restart it 1206762696 M * DanielW does it cleanly umount then . it need not to umount anything correct? 1206762716 M * Bertl it should do fine (well the same way as your real guest does) 1206762766 M * DanielW geting timeout and then it killed the processes of the guest 1206762780 M * Bertl probably same happened on the real host 1206762810 M * DanielW should i now try to start it? 1206762819 M * Bertl yep 1206762861 M * DanielW ok it doesn't start 1206762866 M * Bertl perfect! :) 1206762882 M * DanielW an error occured while executing the vserver startup sequence 1206762901 M * Bertl so, here you go .. this is caused by your rc script somehow 1206762912 M * DanielW it is very likely that the init-script () (why is that empty) failed 1206762973 M * DanielW how can i find out what init script it tries to start? 1206762984 M * Bertl --debug? 1206763142 M * DanielW don't think that there is anything about that in there 1206763158 M * Bertl I'm pretty sure it is .. run it and upload the output :) 1206763412 M * DanielW http://rafb.net/p/N3k5Bq61.html 1206763494 M * Bertl it first looks for /etc/vservers/etch32/vdir//etc/init.d/rc 1206763501 M * Bertl and then for /etc/vservers/etch32/vdir//etc/rc.d/rc 1206763522 M * Bertl note, one of them needs to be there and _executable_ :) 1206763557 M * Bertl the startup runlevel is 3, the stop runlevel 6 1206763569 M * DanielW it needs to be +x ? *G* 1206763574 M * Bertl yep :) 1206763705 M * DanielW ok now it starts :-) 1206763716 M * Bertl so that was simple then :) 1206763733 M * DanielW but ssh in is still not possible . but i think that is no vserver special problem 1206763748 M * Bertl check that the guest is actually running now 1206763754 M * Bertl 'vserver-stat' 1206763759 M * DanielW i can connect to sshd 1206763776 M * DanielW but after entering the correct passwd i get logged out 1206763790 M * Bertl maybe an sshd is running on the host? 1206763806 M * Bertl (if so, you might want to restrict that to host only ips) 1206763840 M * DanielW no it gets connected to the correct machine. the the login in the log file of the guest 1206763856 M * DanielW pam_open_session permission denied after session opend for user root 1206763874 M * Bertl so you can work on fixing that up now too 1206763888 M * DanielW yeah. thanks for your help 1206763918 M * Bertl again, you're welcome! and maybe consider donating something to the Linux-VServer project, if you like it :) 1206764024 M * DanielW yeah i think i like it. but well i think the one how should think of donating something for it, is my hoster he earns money with the vserver install i am using 1206764032 M * DanielW who 1206764051 M * Bertl right, maybe point that out to him then 1206764213 M * DanielW i mean well i am using it. but i am paying someone who gave it to me. well at some point you can think that i am using it this way. but in this setup i think he is the one which should spend. but maybe i will use it here some day and then i may reconsider that (the problem is there are a 100 open source projects who should get donations by me) i prefere to constribute in other ways back to the open source community. 1206764267 M * Bertl but where do you get such a good service? *G* 1206764308 M * DanielW ok you are right that one today was really great 1206764335 M * DanielW i could have found some of them by myself but it would take me a lot more time. so yeah i am very happy about your help :-) 1206764675 M * etteyafed How is the linux-vservers doing financially? Nobody is paying out of pocket for the servers and hosting I hope. 1206764695 M * DanielW good question 1206764719 M * etteyafed I know there are few projects that are doing that from time to time. 1206764751 M * Bertl well, we (the developers) are all paying for the hosting and similar in some or the other way, but most devel machines are somehow contributed 1206764829 M * etteyafed Well I always buy my own devel machines but I know that hosting gets really expensive really fast. 1206764846 M * DanielW Bertl: yeah it is working now remote :-) 1206764854 M * etteyafed Our 10Mbit lines are $300USD alone. 1206764863 M * Bertl DanielW: congrats! 1206764879 M * DanielW now i must get rid of all that unused debian stuff. 1206764907 M * DanielW there is nothing like a remove all files which are not in tar archive? 1206764933 M * Bertl nah, but you could utilize sync with the --delete option :) 1206764938 M * Bertl *rsync 1206765033 M * DanielW don't know how rsync works. never used it :-( i can untar it somewhere and then sync between the two? 1206765049 M * Bertl yep, that's the idea 1206765067 M * Bertl something like rsync -axHPSD --delete 1206765097 M * DanielW but does it delete itself then? will i will just go and read about it 1206765389 M * etteyafed You just paid without exposing your financial information. Receipt ID: XXXXX 1206765410 M * etteyafed information... so valuable 1206765463 M * Bertl thanks a lot! 1206765524 M * DanielW lol now it killed itself. something with libc is going wrong and no program could get started anymore. it tried to find some debian lib. should do the think with this rsync delete from the start 1206765702 M * etteyafed The project we are using the vservers on has the potential to save us quite a bit of headache and prevent a lot of downtime. I can restore a crashed httpd in less than 3 seconds from backed up data, or jusr reboot in about the same time. The hose has almost nothing running on it so it can watch the network services and restart vservers if needed, make on the fly backups (because it's all on the same hw) and restore them in real 1206765781 M * etteyafed If I could only get the DB to work in a vserver... 1206765932 M * Bertl what's the problem with the db? 1206765938 M * etteyafed the selinux thing 1206765951 M * Bertl why do you need selinux for a db? 1206766012 M * etteyafed SE-PostgreSQL, to manage row and cell level access permissions. We use SELinux as what is was meant to be, RBAC, TE, and MLS 1206766105 M * etteyafed It is not really for the system security as a whole, but for access level permissions. I think that if vanilla postgres could do row and cell based permissions mgmt (not just table level) we wouldn't use this other product. 1206766166 M * Bertl so it does user/group checks against SElinux policies, yes? 1206766228 M * etteyafed I don't like being tied to an os (linux) or tech (SELinux). Who knows maybe in another year we want to move to FBSD or OSX but couldn't because we need selinux and linux. 1206766228 M * etteyafed Yes it does the policy checks in the DB. 1206766261 M * etteyafed Its a SELinux enabled (read patched) version of PostgreSQL 1206766773 Q * etteyafed Ping timeout: 480 seconds 1206767799 Q * dowdle Remote host closed the connection 1206768707 M * Bertl okay, off to bed now .. have a good one everyone! 1206768712 N * Bertl Bertl_zZ 1206770953 Q * balbir Ping timeout: 480 seconds 1206771608 J * balbir ~balbir@122.167.210.7 1206772693 J * etteyafed ~etteyafed@68-113-163-245.dhcp.plt.ny.charter.com 1206773679 Q * DanielW Quit: Verlassend 1206774881 M * etteyafed Anyone know how vservers run on fedora 8? 1206774905 M * etteyafed Probably just fine, I just haven't seen a usage case. 1206775552 M * etteyafed How about patching a 2.6.25 kernel to use a vserver? I only see 2.6.22 in downloads. I know that 2.6.25 has a few featurres that might mess with the way vservers work. 1206775648 M * etteyafed Maybe I will try it later. 1206775717 J * dna ~dna@42-220-dsl.kielnet.net 1206775727 M * etteyafed Ahhh... I see how it is done. Well that might be a bit too much work for me ATM. 1206775819 J * xdr ~xdr@137-173-96-87.cust.blixtvik.se 1206776657 Q * dna Quit: Verlassend 1206777836 J * larsivi ~larsivi@144.84-48-50.nextgentel.com 1206780181 J * friendly12345 ~friendly@ppp121-44-224-29.lns2.mel4.internode.on.net 1206781366 J * ftx__ ~ftx@dslb-084-060-234-060.pools.arcor-ip.net 1206782311 M * daniel_hozac 2.6.25 isn't released yet... 1206782599 M * arachnist no? my kernel's at 7.0 already ;) 1206783735 Q * friendly12345 Quit: Leaving. 1206785248 J * esa ~esa@ip-87-238-2-45.static.adsl.cheapnet.it 1206785943 J * hparker ~hparker@208.188.248.81 1206786066 Q * balbir Ping timeout: 480 seconds 1206786751 J * balbir ~balbir@122.167.194.197 1206786912 J * ftx_ ~ftx@dslb-084-060-215-127.pools.arcor-ip.net 1206787298 Q * ftx__ Ping timeout: 480 seconds 1206787502 Q * nenolod Remote host closed the connection 1206787512 J * nenolod ~nenolod@ip70-189-74-69.ok.ok.cox.net 1206787533 Q * nenolod 1206787545 J * nenolod ~nenolod@ip70-189-74-69.ok.ok.cox.net 1206788462 Q * grobie Remote host closed the connection 1206790184 J * Infinito ~argos@200-140-66-14.gnace701.dsl.brasiltelecom.net.br 1206790556 J * tobifix ~tobifix@muedsl-82-207-233-232.citykom.de 1206790744 Q * Infinito Quit: Leaving 1206791176 J * atterdag ~atterdag@c-64c4e455.43-40-64736c10.cust.bredbandsbolaget.se 1206792368 J * hijacker ~Lame@87-126-142-51.btc-net.bg 1206792968 J * dna ~dna@91-241-dsl.kielnet.net 1206793427 J * JonB ~NoSuchUse@77.75.164.169 1206793588 Q * pmjdebruijn Quit: leaving 1206793974 Q * padde Remote host closed the connection 1206794076 Q * bzed Quit: leaving 1206794221 J * bzed ~bzed@devel.recluse.de 1206794242 Q * bzed 1206794249 J * bzed ~bzed@devel.recluse.de 1206794344 J * Julius ~julius@p57B26EF1.dip.t-dialin.net 1206794473 J * padde ~padde@patrick-nagel.net 1206794731 Q * tobifix Remote host closed the connection 1206794941 J * pmjdebruijn ~pmjdebrui@jester.pcode.nl 1206794970 Q * pmjdebruijn 1206794985 J * pmjdebruijn ~pmjdebrui@jester.pcode.nl 1206795273 N * Bertl_zZ Bertl 1206795396 M * Bertl morning folks! 1206795433 J * tobifix ~tobifix@muedsl-82-207-233-232.citykom.de 1206795643 M * nox hi Bertl 1206796062 J * xdr_ ~xdr@137-173-96-87.cust.blixtvik.se 1206796380 Q * xdr Ping timeout: 480 seconds 1206796401 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1206796417 Q * pmjdebruijn Quit: leaving 1206796421 J * pmjdebruijn ~pmjdebrui@jester.pcode.nl 1206796462 Q * pmjdebruijn 1206796472 Q * tobifix Remote host closed the connection 1206796589 J * pmjdebruijn ~pmjdebrui@jester.pcode.nl 1206796703 Q * tokkee Ping timeout: 480 seconds 1206796919 J * tokkee tokkee@ssh.faui2k3.org 1206797334 J * Infinito ~argos@200-140-66-14.gnace701.dsl.brasiltelecom.net.br 1206797453 Q * harry Ping timeout: 480 seconds 1206798764 J * pmenier ~pmenier@ACaen-152-1-96-236.w86-215.abo.wanadoo.fr 1206799020 Q * Aiken Remote host closed the connection 1206799110 M * pmenier Hello 1206799127 M * pmenier Bertl: ping ? 1206799157 M * pmenier http://paste.linux-vserver.org/11892 1206799275 M * Bertl could you upload your .config somewhere? 1206799353 M * pmenier yes 1206799415 M * pmenier http://www.pmenier.net/pb/config26244 1206800572 M * Bertl please add #include before the vs_device.h include 1206800600 M * pmenier ok 1206800772 J * dowdle ~dowdle@67-42-224-198.blng.qwest.net 1206800867 M * Bertl pmenier: you also need to disable ocfs2 1206800882 M * pmenier ah... 1206800882 M * Bertl (it's not compiling yet) 1206800910 M * pmenier i stop compil and correct 1206801248 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1206801450 Q * JonB Quit: This computer has gone to sleep 1206801587 M * pmenier seems good CC fs/char_dev.o => no error 1206802001 J * Piet ~piet@tor.noreply.org 1206802843 Q * Piet Remote host closed the connection 1206802894 J * Piet ~piet@tor.noreply.org 1206803573 M * zbyniu Bertl: http://zz.iapt.pl/vserver-ocfs2-mount_opt.patch not tested but hits proper structs and builds now 1206803712 M * pmenier Bertl: build complete . 2 vservers started 1206803869 M * Bertl zbyniu: ah, nice, will check shortly 1206803915 M * zbyniu :) 1206804164 J * Slydder ~chuck@dslb-088-072-013-203.pools.arcor-ip.net 1206805236 J * mrfree ~mrfree@host174-177-dynamic.32-79-r.retail.telecomitalia.it 1206806072 J * julius_ ~julius@p57B2771B.dip.t-dialin.net 1206806525 Q * Julius Ping timeout: 480 seconds 1206806583 Q * nkukard Ping timeout: 480 seconds 1206806633 Q * Piet Ping timeout: 480 seconds 1206806646 M * Bertl zbyniu: looks good! thanks, will be included in the next version 1206806897 J * nkukard ~nkukard@196.212.73.74 1206807255 J * Piet ~piet@tor.noreply.org 1206807748 Q * balbir Ping timeout: 480 seconds 1206807765 J * virtuoso_ ~s0t0na@ppp89-110-58-120.pppoe.avangarddsl.ru 1206807814 Q * virtuoso Read error: Connection reset by peer 1206808405 J * balbir ~balbir@122.167.212.6 1206808939 J * JonB ~NoSuchUse@77.75.164.169 1206810374 M * etteyafed daniel_hozac: I know 2.6.25 is still in rc but it will soon be out. I was just wondering how much work is involved in porting vservers from one kernel version to the next, and how one might go about doing it. And where to start. I don't want to just try and apply the patch, because there is a new scheduler and support for "dyn-ticks" not exactly sure what that means. 1206810395 M * etteyafed But I could guess that it might interfere with the way vserver works. 1206811390 Q * Slydder Quit: Leaving. 1206811617 M * pmjdebruijn etteyafed: dynticks are also in .24 1206811637 M * pmjdebruijn etteyafed: and vserver support for .24 isn't done yet 1206811748 M * etteyafed What kind on time frame are you looknig at for 2.6.24? I might be able to sub 24 for 25 until there is support for 25 1206811757 M * etteyafed I don't need 25 1206811865 M * pmjdebruijn might be able? 1206811872 M * pmjdebruijn what does 25 have that you need? 1206811885 M * etteyafed Nothing. 1206811886 M * pmjdebruijn there are beta's of 24 1206811891 M * etteyafed I would be able 1206811892 M * pmjdebruijn but it might still take a while 1206811926 M * pmjdebruijn I'm still at 22 1206812046 M * etteyafed Betas you say? I am using .22 now but there is a machine that we have that runs software that "depends" on a .24 kernel. It might actually work on .22 I haven't tested it, but the docs say that it needs .24. I would like to get rid of that machine if I could though, and migrate that stuff to the bigger server that hosts most of our vservers. 1206812093 Q * quasisane Remote host closed the connection 1206812121 M * etteyafed The server room is a lot crowded and we run into power problems now and then. The few physical machines we have the better. And none of the systems are over 30% load right now. 1206812195 M * etteyafed I would like to just have 3-4 physical servers and another 3-4 on hot standby. 1206812263 J * tobifix ~tobifix@muedsl-82-207-213-095.citykom.de 1206812377 Q * hijacker Quit: Leaving 1206812414 M * pmjdebruijn etteyafed: what software depends on .24 then? there might be others ways to solve it... 1206812437 J * virtuoso ~s0t0na@ppp91-122-59-191.pppoe.avangarddsl.ru 1206812849 Q * virtuoso_ Ping timeout: 480 seconds 1206813027 N * pmenier pmenier_off 1206815291 J * mick_work ~clamwin@h-74-2-196-226.miatflad.covad.net 1206815929 N * DoberMann[PullA] DoberMann 1206816479 J * Beuc ~yo@82.238.35.175 1206816558 M * Beuc Hi. I was wondering if a 2.6.24 release is planned :) Does this present difficulties? 1206816612 M * Bertl there is a devel release for 2.6.24.4 1206816632 M * Bertl and yes, mainline changes caused a lot of things to break 1206816672 M * Beuc OK, thanks :) 1206816764 J * ntrs ~ntrs@77.29.65.116 1206817089 J * Xilaw ~myster0@trinity.matrix-world.org 1206817201 P * Beuc Leaving 1206817220 Q * ftx_ Remote host closed the connection 1206817946 Q * Infinito Quit: Leaving 1206818465 J * jpacheco ~justin@CPE000e7b691297-CM0012c9a00ac8.cpe.net.cable.rogers.com 1206818478 M * jpacheco hola everyone 1206818489 M * jpacheco just a quick question 1206818507 M * jpacheco has vserver development stopped for the latest kernels?? 1206818939 M * Bertl why? 1206819801 M * jpacheco just haven't seen a patch against the latest kernels for a while 1206819864 M * Bertl well, look again, preferable here: http://vserver.13thfloor.at/Experimental/ 1206819939 M * jpacheco are the patches here stable or in dev? 1206819949 Q * tobifix Remote host closed the connection 1206819964 M * Bertl devel patches of course, stable for recent mainline kernels will take a lot of testing 1206820131 M * jpacheco ic 1206820134 M * jpacheco thanks for the update 1206820139 M * Bertl you're welcome! 1206820142 M * jpacheco hows the project goning btw? 1206820147 M * jpacheco still going strong? 1206820173 M * Bertl well, not as much time available as I would like, and almost no funding ... 1206820191 M * Bertl (not many donations either) 1206820211 M * jpacheco that's a shame. its such a good product 1206820246 M * Bertl folks not always go for the good and technical superior :) 1206821313 Q * atterdag Ping timeout: 480 seconds 1206824581 M * etteyafed Bertl: So, on that patch-2.6.24.4-vs2.3.0.34.diff... I should apply that to the vanilla 2.6.24? Have you used that kernel much? Is it reasonably stable? 1206824632 M * daniel_hozac uh, you realize it's the devel tree for an incomplete port? 1206824633 M * Bertl nah, to 2.6.24.4 1206824663 M * Bertl and no, I have no idea about 2.6.24.4 and how stable it is :) 1206824688 M * daniel_hozac Bertl: are any of the known problems fixed? 1206824704 A * daniel_hozac haven't really been keeping track lately. 1206824705 M * etteyafed Yeah I realize that it is in devel for a reason, but I just wanted to do some testing with it. I do a lot of testing you know. 1206824759 M * daniel_hozac you misunderstand. devel is fine, i've been running it in production for months. it's the 2.6.24 part that's untested/unknown/broken/etc. 1206824764 M * Bertl daniel_hozac: nothing of the basic issues have been fixed, but I went over some of the minor details and mended them 1206824781 M * daniel_hozac Bertl: okay. hopefully i'll have some time to spare soon... 1206824868 M * Bertl yeah, after all, we'll get the 'build machine' tomorrow or so, that should help a little with compiling and testing 1206824926 M * etteyafed What is this VServer Control Daemon. Sounds very interesting. 1206824963 M * Bertl I guess Hollow has details on that 1206824983 M * ard eh 1206824989 A * ard dances 1206824998 M * ard a 2.3.0 for 2.6.24 to test :-) 1206825019 M * Bertl go for it .. let us know how it goes 1206825035 M * ard first have to check the grub versions :-) 1206825048 M * ard older grubs can't boot kernels >= 2.6.23 1206825121 M * etteyafed ard: let me know how that goes. 1206825146 M * etteyafed and possibly details. Please. 1206825220 M * daniel_hozac who would be running a grub from 2002 with a really new kernel? 1206825226 M * ntrs daniel_hozac, I have a problem with vhashify 1206825235 M * daniel_hozac ntrs: oh? 1206825254 M * ard well..., upgrading grub in debian fortunately never ever upgrades the grub on the /boot .. 1206825311 M * ntrs It is not creating the links it seems when I do the buld with clone option 1206825351 M * ard are you sure the source is already hashified? 1206825410 M * ntrs Yes I am sure 1206825434 M * Hollow etteyafed: it is a daemon for controlling vservers via rpc 1206825470 M * Hollow etteyafed: but it has been tested very little and i've started to rewrite the rpc implementation, though i have little time for it currently 1206825521 M * ard ntrs : if you showattr the source file, will it show up as cow? 1206825591 M * ntrs example: 1206825594 M * ntrs # showattr /vservers/0000centos5/bin/bash 1206825594 M * ntrs ----UIX /vservers/0000centos5/bin/bash 1206825602 M * ntrs is UIX what I need 1206825607 M * daniel_hozac yes. 1206825648 M * daniel_hozac how are you building your guests? 1206825674 M * ntrs vserver test2 build -m clone -- --source 0000centos5 1206825715 M * daniel_hozac and then showattr /vservers/test2/bin/bash; ls -li /vservers/test2/bin/bash shows what? 1206825743 M * ntrs There is no such file 1206825771 M * daniel_hozac uh, what files are there? 1206825775 M * ntrs For some reason the files from /etc/vservers/0000centos5 are in both /etc/vservers/test2 and /vserves/test2 1206825789 M * ard Hmmmm.... clone in another directory 1206825819 M * ntrs ard, what do you mean? 1206825824 M * ard I found out that if you are in /etc/vservers, it will copy the directory with that name, instead of something from /vservers 1206825825 M * daniel_hozac i assume you were in /etc/vservers when you ran vserver ... build? 1206825846 M * ntrs daniel_hozac, I am not sure 1206825857 M * ntrs Where do I need to be when cloning? 1206825881 M * ard so clone copies the directory in your current directory if it can find one, and only if it can find none, it will really clone /vservers/ 1206825897 M * ntrs ard, wow, I had no idea. 1206825904 M * Bertl s/can/can't/ sounds interesting :) 1206825908 M * ard so anywhere that doesn't have a directory with that name :-) 1206825938 M * Bertl if that's true, it'S probably the first 0.30.215 bug report :) 1206825938 M * ard anyway, that's what it looks like... :-) 1206825939 M * ntrs daniel_hozac, is there a way to clone a non-started guest/template? 1206825947 M * daniel_hozac hmm? 1206825966 M * Bertl daniel_hozac: I guess ntrs means hashify 1206825966 M * daniel_hozac you should avoid cloning running guests. 1206825971 M * ard Bertl : well, I was actually not sure if it was a bug, or a feature :-) 1206825994 M * ntrs daniel_hozac, yes, hashify, sorry. 1206826011 M * ard Bertl : if I am correct, it behaves like -m copy 1206826044 M * daniel_hozac ntrs: not at the moment. 0.30.216 should allow that though. 1206826064 M * ard hmmm, no copy, so it really is something with clone 1206826164 M * ntrs ok, I tried that and /vservers/test2 is not created 1206826178 M * daniel_hozac hmm? 1206826184 M * ard tried what? 1206826216 M * ard you must first remove test2 before cloning again :-) 1206826250 M * ntrs ok, I tried with test5 1206826257 M * ntrs vserver test5 build -m clone -- --source 0000centos5 1206826273 M * ntrs nothing in both /etc/vservers and /vservers 1206826288 M * ard hmmm 1206826297 M * daniel_hozac no output at all? 1206826312 M * daniel_hozac run with --debug 1206826557 M * ntrs http://paste.linux-vserver.org/11894 1206826645 M * daniel_hozac those errors are why it's failing. 1206826662 M * ntrs What errors? can you please help me resolve the problem? 1206826690 M * daniel_hozac line 506 and 507. 1206826727 M * Bertl could that be broken symlinks? 1206826768 M * ntrs I think they are 1206826777 M * ntrs should i just remove them or recreate them? 1206826783 J * Aiken ~james@ppp121-45-192-61.lns1.bne1.internode.on.net 1206826797 M * Bertl I think a broken symlink should not affect the clone procedure 1206826805 M * ard /etc/vservers/.defaults/vdirbase/ should point to /vservers probably 1206826806 M * Bertl but it might be the reason for this failure 1206826836 M * daniel_hozac i can't reproduce it at least. 1206826851 M * daniel_hozac what is /dev/reboot in centos5? 1206826853 M * Bertl let's check the attributes on them 1206826873 M * Bertl the template/guest was hashified, maybe that did something to them 1206826874 M * ntrs I removed them because I don't think I need them and it seems it worked. 1206826890 M * ntrs I can see the whole directory structure of the template in /vservers/test5 1206827613 M * Bertl nap attack ... maybe back later ... 1206827628 N * Bertl Bertl_zZ 1206828253 M * ard f00d attack 1206828298 J * ntrs_ ~ntrs@77.29.66.207 1206828308 M * ntrs_ Do I need to have "/etc/vservers/vserver-name/apps/vunify/hash" with id and method files in each guest for the vhashify to work? 1206828331 M * ntrs_ right now I only have "/etc/vservers/vserver-name/apps/vunify/" and nothing in there 1206828339 M * daniel_hozac that's sufficient. 1206828359 M * ntrs_ how does it know what to hashify with? what the template is? 1206828378 M * daniel_hozac there's no such thing. 1206828388 M * ntrs_ or, is everything compared with the main hash database? 1206828395 M * daniel_hozac /etc/vservers/.defaults/apps/vunify has the required entries. 1206828412 M * ntrs_ for all guests, all templates, all distros? 1206828418 M * daniel_hozac yes. 1206828428 M * ntrs_ Wow, that must be one huge mess :) 1206828639 M * ntrs_ du -csx /vserver/test5 shows about 1.1 GB, if I enter the guest df shows about 120 MB. Can you tell me more about how disk space accounting works here? 1206828667 M * daniel_hozac run du -csx /vservers/.hash /vservers/test5 1206828671 M * ntrs_ I mean what is the correct way to determine disk usage inside and outside of the guest? 1206828687 M * daniel_hozac vdu 1206828732 M * ntrs_ ok, what about the user that is inside the guest? Isn't is weird for him to see only 120 MB used on a template with 1.5 GB worth of files? 1206828742 Q * ntrs Ping timeout: 480 seconds 1206828750 M * daniel_hozac i don't think they'll be complaining ;) 1206828790 M * ntrs_ I agree, but the number is incorrect and someone might ask so we have to know what to tell him/her 1206828802 M * daniel_hozac it's not incorrect. 1206828810 M * daniel_hozac df is counting the files that belongs to the guest. 1206828836 M * ntrs_ yes, it shows about 120 MB but in reality there are 1.5 GB 1206828848 M * daniel_hozac so 1.4 GiB of that is shared. 1206828902 M * ntrs_ Ok, do you think that the end user should have that number? 1206828931 M * daniel_hozac sure. 1206829010 M * ntrs_ I think the end user should not know anything about the outside world, and should see that his guest is using 1.5 GB of space. 1206829077 M * daniel_hozac feel free to sponsor that... 1206829077 Q * JonB Quit: This computer has gone to sleep 1206830970 J * atterdag ~atterdag@c-64c4e455.43-40-64736c10.cust.bredbandsbolaget.se 1206831443 Q * ntrs_ Ping timeout: 480 seconds 1206831948 N * DoberMann DoberMann[ZZZzzz] 1206832008 Q * etteyafed Ping timeout: 480 seconds 1206832403 Q * atterdag Quit: Leaving 1206832664 Q * fatgoose_ Quit: fatgoose_ 1206832986 J * hparker ~hparker@208.188.248.81 1206832993 J * ktwilight ~ktwilight@192.84-66-87.adsl-dyn.isp.belgacom.be 1206833213 Q * ktwilight_ Ping timeout: 480 seconds 1206833845 Q * dna Quit: Verlassend 1206834891 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1206834893 J * friendly12345 ~friendly@ppp121-44-224-29.lns2.mel4.internode.on.net