1136508264 Q * entroposcope Ping timeout: 480 seconds 1136510172 N * lchvdlch lchvdlog 1136510174 M * lchvdlog g'nite all 1136510183 M * Bertl night lchvdlog! 1136510671 P * undefined 1136511722 J * koglinjg ~koglinjg@eris.osd.wednet.edu 1136511730 M * koglinjg Hello 1136511760 M * koglinjg I am trying to get vserver to work on a gentoo box with prtage. Can anyone help me with this? 1136511771 M * koglinjg portage* 1136511874 M * Bertl welcome koglinjg! 1136511890 A * Bertl is no gentoo guy, bu he will try :) 1136511940 M * koglinjg That is all that i am asking is for someone to try to help if they can 1136511953 M * koglinjg its actually the vunify script that i need to hack 1136511967 M * koglinjg its is written for a rpm based system 1136511998 M * Bertl ahh, well, that is probably simple ... have a look at vhashify 1136512017 M * Bertl (it basically replaces vunify) 1136512036 M * koglinjg i will check that on google and i will be back in a bit 1136512060 M * Bertl np, feel free to hang around 1136512272 M * koglinjg well i check it out and it looks good in the test run in my head so now i am doing it so thank you very much for that 1136512284 M * Bertl you're welcome! 1136512297 M * koglinjg I will hang around for a bit untill i have completely this successfully or until it starts working on data 1136512319 M * Bertl feel free to hang around even longer :) 1136512476 M * koglinjg can i send you a link 1136512482 M * koglinjg to show ya what i am looking at? 1136512557 M * Bertl sure, just paste it here, or if it is sensible data, /msg me 1136512569 M * koglinjg  1136512575 M * koglinjg http://linux-vserver.org/alpha+util-vserver 1136512590 M * koglinjg there is a section on vhashify about 3/4 down 1136512641 M * koglinjg "in each used filesystem" so filesystem refers to each vserver 1136512663 M * Bertl well, depends, it really means 'filesystem' here 1136512678 M * Bertl so, if all your guests are on one filesystem, then only one 1136512693 M * koglinjg so filesystem would be each partition? 1136512706 M * Bertl if OTOH, your guests are on different filesystems, e.g. each on one, then this would mean for each 1136512727 M * Bertl koglinjg: filesystem, partitions just carry filesystems sometimes :) 1136512736 M * koglinjg ok thanks 1136512743 M * koglinjg so in my case just one 1136512753 M * Bertl e.g. partition could be an lvm partition with many filesystems on it 1136512865 Q * infowolfe_ Read error: Connection reset by peer 1136512904 M * koglinjg so when i run vserver hashify "can not determine packagemanagment style 1136512915 M * koglinjg "failed to determine config files" 1136513028 M * Bertl on gentoo (no packages the tools understand yet) you probably have to go for the --manual 1136513041 M * Bertl *--manually 1136513056 M * Bertl and specify some pathes like /sbin and /bin 1136513087 M * Bertl but it might be easy to add them to some config file (let me see if I can find some links) 1136513150 M * koglinjg so vhasify --manually /sbin /bin (whats the syntax? for multiple dirs) 1136513163 M * koglinjg vhashify*** 1136513281 Q * aj__ Ping timeout: 480 seconds 1136513863 M * Bertl koglinjg: you specify the path to the guest, and an exclude list 1136513878 M * Bertl koglinjg: what kernel version/patches do you use? 1136513905 M * koglinjg sry 1136513910 M * Bertl (just had to fix powerpc :) 1136513911 M * koglinjg i wsnt there for a sec 1136513929 M * koglinjg umm what do i put in the exclude list? 1136513942 M * koglinjg does it assume do all except the exclude list 1136513944 M * Bertl depends on your kernel patch version, with devel kernel 1136513961 M * Bertl you do not need to put anything there, if you enabled the CoW link breaking 1136513982 M * koglinjg 2.6.13.3 -vs2.1.0-rc3-gentoo 1136513994 M * Bertl without CoW or stable branch you want to exclude files which get written or appended to 1136514007 M * Bertl like those in /etc or /var 1136514042 M * Bertl koglinjg: you know how unification works? 1136514060 M * koglinjg it uses hard links 1136514069 M * Bertl yes, immutable but unlinkable ones 1136514087 M * koglinjg and therefore 1136514246 M * Bertl therefore you want to avoid unifying files which get appended or written to (except when you have CoW enabled) 1136514605 M * koglinjg what foramt do the exclude lists need to be in 1136514620 M * koglinjg enter seperated directories to exclude? 1136514729 M * Bertl good question, I'd say a path per line 1136514738 M * koglinjg "could not find the place for hashified files" 1136514748 M * koglinjg so i think it was ok with the exclude file 1136514754 M * koglinjg yay 1 down 1136514765 M * Bertl probably vhashify also honors the exclude files configured in apps/vunify 1136514851 M * koglinjg yayayaya i think we got it 1136514854 M * koglinjg :) 1136514875 M * Bertl excellent! maybe add some notes to the wiki? 1136514883 A * koglinjg tosses koglinjg a cookie... good boy 1136514895 M * Bertl lol 1136514896 M * koglinjg lol oh toss cokie to not from 1136514927 A * koglinjg tosses Bertl a cookie... good boy 1136514963 M * Bertl tx for the cookie! 1136515031 M * koglinjg as u can tell this is my first time on a command line irc client 1136515062 J * entroposcope ~entroposc@user-0c992og.cable.mindspring.com 1136516063 M * Bertl welcome entroposcope! 1136516606 Q * koglinjg Quit: gtkBitchX-1.1-final -- just do it. 1136518892 J * stefani ~stefani@c-24-19-46-211.hsd1.wa.comcast.net 1136519228 M * Bertl wb stefani! 1136519244 M * stefani hola 1136519602 M * Bertl stefani: btw, what machines do you use for linux-vserver (if any at all)? 1136519613 M * stefani several 1136519630 M * stefani at least 5 1136519632 M * Bertl mainly x86 or x86_64 arch? 1136519657 M * stefani one hosts a dozen guests, another does a dozen or so. x86 1136519666 M * stefani i have no x86_64 et al. 1136519677 M * stefani no SMP either. 1136519694 M * Bertl okay, so x86/UP fits the main usage profile ... 1136519728 M * stefani money. 1136520000 Q * Aiken_ Quit: Leaving 1136520062 M * Bertl stefani: what will come first, the move towards SMP or 64bit? 1136520084 M * stefani welll i see more people doing 64 now. 1136520104 M * stefani previously it was SMP. but now more 64, and of course SMP-64 too 1136520104 M * daniel_hozac i think they will come simuletaneously. 1136520118 M * daniel_hozac AMD64x2 1136520145 M * stefani i helped install an AMD64x2 last week. 1136520150 M * Bertl well, with the upcoming intel dual core chips, the 64bit might have to wait, no? 1136520152 M * stefani not vserver, but 1136520164 M * daniel_hozac aren't the Intel dual core chips 64-bit too? 1136520178 M * stefani the XEONS act like 64 1136520186 M * Bertl I'm pretty sure there will be a cheap version too 1136520188 M * daniel_hozac EM64T or whatever Intel is calling it. 1136520198 M * daniel_hozac well, the cheap version is the non-dual core one ;) 1136520221 M * Bertl yes, emt64 is the intel version of amd64 (aka linux x86_64) 1136521490 M * mnemoc afaik emt64 is considered an optimization for x86_64 arch 1136521698 M * Bertl hmm, by whom? 1136521744 M * mnemoc gcc iirc 1136521757 M * mnemoc but of course i can be wrong 1136521759 M * daniel_hozac probably just instruction ordering. 1136521803 Q * shuri Remote host closed the connection 1136521854 J * balbir ~balbir@59.92.144.226 1136521949 M * Bertl mnemoc: ah, well, I buy it that gcc supports amd64 and emt64 as 'special' subarchs of x86_64 ... 1136521964 M * mnemoc :) 1136521967 M * mnemoc thanks 1136522072 M * Bertl just built and started the first debian guest on a quad power5 :) 1136522089 M * daniel_hozac nice! 1136522116 M * Bertl and guess what, with the debian util-vserver package!! 1136522157 M * daniel_hozac haha, wow. 1136522173 M * Bertl I see a light at the end of the tunnel ... :) 1136522191 M * Bertl .o( incoming train? :) 1136522203 M * mnemoc :) 1136522302 M * Bertl does anybody have some good links for vDSO? 1136522833 M * Bertl wasn't somebody working on a test script for disk limits/quota? 1136522854 M * Bertl is this already done? 1136523303 M * Bertl Cpu0 : 0.0% us, 0.0% sy, 24.6% ni, 75.4% id, 0.0% wa, 0.0% hi, 0.0% si 1136523304 M * Bertl Cpu1 : 0.0% us, 0.0% sy, 24.5% ni, 75.5% id, 0.0% wa, 0.0% hi, 0.0% si 1136523304 M * Bertl Cpu2 : 0.0% us, 0.1% sy, 24.3% ni, 75.5% id, 0.1% wa, 0.0% hi, 0.0% si 1136523304 M * Bertl Cpu3 : 0.0% us, 0.0% sy, 24.4% ni, 75.6% id, 0.0% wa, 0.0% hi, 0.0% si 1136523317 A * Bertl really loves the new per-cpu scheduler :) 1136523445 M * mnemoc :) 1136523633 M * Bertl okay, enough for me for tonight ... after all, a new arch finished and tested ... 1136523642 M * Bertl have fun, and cya all tomorrow! 1136523650 N * Bertl Bertl_zZ 1136527295 M * eyck woa, what per-cpu scheduler? 1136528638 P * stefani parting (is such sweet sorrow) 1136529859 J * meandtheshell ~markus@85-124-10-194.dynamic.xdsl-line.inode.at 1136530227 J * infowolfe jthm@209-112-210-87-cdsl-rb1.nwc.acsalaska.net 1136530248 Q * balbir Quit: Leaving 1136531341 J * tudenbart ~willi@xdsl-81-173-169-192.netcologne.de 1136531766 Q * dothebart Ping timeout: 481 seconds 1136533083 J * schellh ~bla@ipsio389.ipsi.fraunhofer.de 1136533089 M * schellh hello 1136533206 M * schellh i got a problem with vserver 2.01 sparc: trying to build a debian guest 1136533230 M * schellh E: Invalid Release file, no entry for main/binary-TI UltraSparc II (BlackBird)/Packages 1136533910 M * schellh within debootstrap_0.1.17.7woody1_sparc.deb 1136536249 M * schellh *sniff 1136536822 J * Milf ~Miranda@ipsio184.ipsi.fraunhofer.de 1136536836 M * Milf 'morning Folks 1136537099 Q * shedi Quit: Leaving 1136537106 M * schellh hi you suckerrr 1136537128 M * Milf Don't mind him, he sits two meters across from me. 1136537136 M * Milf and is perpetually bored :) 1136540303 J * prae ~prae@ezoffice.mandriva.com 1136540535 M * schellh hello 1136541085 J * shedi ~siggi@tolvudeild-203.lhi.is 1136541415 J * Johan ~kvirc@194.151.113.2 1136541424 M * Johan Hi all 1136541554 M * Johan I have a production machine with linux-2.4.28 and vs-1.29. The machine has a high load, sometimes very high, without actually doing anything. I wonder if this perhaps might be a leak or configuration problem with the vserver patch I use (2.428 with vs 1.29). Perhaps anyone an idea what to look for? 1136541741 M * TheSeer run vtop on the host to actually see what process and maybe vserver causes the load 1136541765 M * TheSeer the vserver "patch" can't have anything to do with it per se, since it basically doesn't do anything itself 1136541788 M * TheSeer a vserver is nothing but a bunch of seperated processes 1136541801 M * Johan Ok thanx. I will try vtop 1136541817 M * TheSeer so, simply speaking, if the last process within a context dies, the vserver is gone 1136541829 M * TheSeer (i know that's over simplified, but shows the point) 1136541904 M * Johan I understand, thanx. Wathcing vtop now 1136542205 M * schellh anyone in here familiar with vserver -sparc ? 1136542227 M * schellh E: Invalid Release file, no entry for main/binary-TI UltraSparc II (BlackBird)/Packages - while debootstrap_0.1.17.7woody1_sparc.deb bootstrapping 1136545816 J * balbir ~balbir@59.92.156.132 1136545840 Q * balbir Quit: 1136545841 M * FaUl schellh: works for me 1136545850 M * FaUl schellh: try sarge :-) 1136545864 M * FaUl schellh: woody on alpha is broken as well ;-) 1136545975 J * Smutje ~Smutje@xdsl-87-78-6-166.netcologne.de 1136546081 Q * Smutje_ Ping timeout: 480 seconds 1136546321 J * sod ~sod@balls.uniques.ro 1136546611 M * sod helo, let me be a fool for a minute and ask you smth; I've been using vserver for quite a while now on many servers but now I am faced with a very strange issue; let's say we have a ssh daemon in a vserver. If I telnet from anywhere to the vserver ip on port 22 it works just fine. If I do vserver vserver_name enter and try to telnet on the same ip port 22 it just sits there. Any ideas? 1136546742 M * TheSeer so to sum it up, you try to login to service of the vserver from within the vserver 1136546759 M * sod yes, ssh is just an example 1136546778 M * sod let me paste it to you 1136546780 M * sod # ip a s 1136546780 M * sod 1: eth0: mtu 1500 qdisc pfifo_fast qlen 1000 1136546780 M * sod link/ether 00:03:1d:01:e3:50 brd ff:ff:ff:ff:ff:ff 1136546780 M * sod inet 80.86.106.18/24 brd 80.255.255.255 scope global eth0 1136546780 M * sod inet 83.166.201.225/24 brd 80.255.255.255 scope global eth0:ines01 1136546796 M * sod telnet 83.166.201.225 22 1136546796 M * sod Trying 83.166.201.225... 1136546796 M * sod Connected to 83.166.201.225. 1136546796 M * sod Escape character is '^]'. 1136546796 M * sod SSH-2.0-OpenSSH_3.8.1p1 Debian-8.sarge.4 1136546804 M * sod this was from the outside 1136546810 M * sod now I enter the vserver 1136546832 M * sod ipv4root is now 83.166.201.225 1136546832 M * sod New security context is 101 1136546832 M * sod ines01:/# telnet 83.166.201.225 22 1136546832 M * sod Trying 83.166.201.225... 1136546832 M * sod 1136546840 M * sod and just sits there 1136546877 M * TheSeer do a netsat -a within the vserver 1136546893 M * TheSeer netstat -an | grep 22 1136546927 M * sod it's ok: 1136546928 M * sod tcp 0 0 83.166.201.225:22 0.0.0.0:* LISTEN 19146/sshd 1136546950 M * TheSeer true.. 1136546955 M * TheSeer interesting ;> 1136546978 M * TheSeer what versions of things do you have? 1136546992 M * sod 2.6.14.3-vs2.0.1 1136547001 M * TheSeer hmm 1136547005 M * TheSeer let me check on my servers 1136547006 M * sod util-vserver 0.30.204-5sarg 1136547109 M * sod weird thing is that even tcpdump is silent 1136547120 M * TheSeer works for me with smtp 1136547127 M * sod it should work, on any other servers works just fine 1136547127 M * TheSeer i don't have a vserver running ssh 1136547135 M * sod no, ssh is just an exemple 1136547139 M * sod example 1136547149 M * TheSeer i do realize that ;) 1136547152 M * sod it's the same for anything, even netcat 1136547163 M * sod hmm, I'll try to downgrade 1136547164 M * TheSeer i just tried on a server running an smtpd 1136547176 M * TheSeer telneting to it works from within the vserver 1136547183 M * TheSeer 2.6.14.4-vs.2.1.0 here 1136547306 M * sod all my other servers run 2.6.9-vs1.9.3 1136547313 M * sod think I'll go back to that for this one 1136547335 M * TheSeer if that fixes it, please try 2.6.15-vs2.1.0 1136547340 M * TheSeer and see if that works 1136547349 M * TheSeer otherwise talk to Bertl and report a bug ;p 1136547369 M * sod yeap 1136547383 M * sod btw, there is another strange thing with this 2.x release 1136547440 M * sod for the older ones, everithing worked just fine, for 2.x you have to select advanced router features in the kernel config to have networking in vservers 1136547455 M * sod I think it's somewhat related 1136547533 M * TheSeer hmm.. you might ask Bertl about that too... 1136547541 M * TheSeer i just copied my .config over 1136547542 M * TheSeer ;) 1136547582 M * sod bertl seems to be sleeping 1136547605 M * sod and one thing he will most likely tell me is to switch to the new style configs 1136547631 M * sod which I find absolutely difficult to maintain 1136547674 M * sod we invested to much time in creating an automativ system for deploing vservers tu rewrite most of it now :( 1136547866 M * TheSeer actually i think the new style config to be easier to handle 1136547876 M * TheSeer since you can just drop in files or remove them 1136547880 M * TheSeer rather then parsing complex files 1136547904 M * TheSeer but i agree it's a pretty heavy change ;) 1136547937 M * sod I busted their asses to write me a system to parse and create all that 1136547956 M * sod I really think they will hurt me is I tell them they could just create files :)) 1136548330 M * Johan TheSeer: I am still strugling with my high-load. Simple question: If almost every proces uses 0.0% cpu and mem, vtop shows the system is +/- 99.4% idle all the time, how can the load be around 2.40 or 3.00 ? 1136548374 M * sod system load; bad drivers, bad IO subsystems, crappy network cards, etc 1136548394 M * TheSeer vtop should display that... 1136548404 M * TheSeer hmm.. 1136548469 M * Johan Ok, can it also be a high-mem problem? The machine has 4gb memory (sorry, I know this is not vserver specific anymore) 1136548526 M * sod is it ia32 or x86-64 ? 1136548550 M * sod iae, shouldn't be a highmem issue 1136548567 M * Johan Intel Xeon 1136548594 M * sod xeon comes with an EMT64 version as well 1136548622 M * sod if you have it I strongly encourage you tu switch to a 64 bit kernel and software 1136548751 M * TheSeer Xeon as EMT64? 1136548761 J * oliwel ~mail-at-o@host-62-245-151-178.customer.m-online.net 1136548897 M * ElPistoleroRoey Bertl_zZ: my question was, how hard is it to get OpenVPN running on VServer? 1136548899 M * ElPistoleroRoey TheSeer: yes 1136548906 M * sod yeap, I have several 1U machines with 2.8 xeons an running with emt64 1136548943 M * TheSeer sod: how do i tell if the xeon "my" server has is 64bit? ;> 1136548966 M * TheSeer ElPistoleroRoey: you'd need nget for that if you want to it to be secure 1136548973 M * sod uhh, tough one 1136548975 M * ElPistoleroRoey hmm 1136548976 M * ElPistoleroRoey nget?? 1136548992 M * ElPistoleroRoey TheSeer, sod: actually it's listed on one of the dev's TODO list 1136548997 M * ElPistoleroRoey to get OpenVPN running in vserver. 1136549001 M * TheSeer next generation networking 1136549008 M * TheSeer ngnet 1136549016 M * ElPistoleroRoey As OpenVPN needs to be able to open tunN devices. 1136549025 M * TheSeer it's more then just the device 1136549047 M * TheSeer you'd need to have vserver specific routing information i'd say 1136549053 M * ElPistoleroRoey oh 1136549055 M * ElPistoleroRoey hmm right right. 1136549065 M * ElPistoleroRoey maybe it's something that vserver needs to make specific accommodations for 1136549066 M * ElPistoleroRoey ? 1136549068 M * TheSeer and what happens if you modify the default gw for the vserver? 1136549076 M * ElPistoleroRoey (dunno) 1136549079 M * TheSeer ;) 1136549093 M * TheSeer i'd say that it won't be possible until ngnet is done 1136549168 M * ElPistoleroRoey ok 1136549173 M * ElPistoleroRoey is ngnet part of Linux or is it OpenVPN? 1136549197 M * TheSeer it's a vserver thing 1136549214 M * TheSeer and i'm not sure as to how much of that is already done in code 1136549627 M * TheSeer Recognizing IA64 Enabled CPU (Useful for Xeon(tm) Processors) 1136549627 M * TheSeer cat /proc/cpuinfo 1136549627 M * TheSeer Check for the 'lm' flag. 1136549629 M * TheSeer found it ;> 1136549640 M * TheSeer so i don't have 64bit on that box.. to bad ;> 1136550070 J * lilalinux ~plasma@dslb-084-058-203-051.pools.arcor-ip.net 1136550148 Q * lilalinux Quit: 1136550153 J * lilalinux ~plasma@dslb-084-058-203-051.pools.arcor-ip.net 1136550795 M * ElPistoleroRoey http://linux-vserver.org/Todo+List 1136550797 M * ElPistoleroRoey there. 1136550799 M * ElPistoleroRoey it's there. 1136551407 Q * oliwel Quit: Chatzilla 0.9.69.1 [Firefox 1.5/2005111116] 1136551578 J * khnz ~chatzilla@shaka.zrz.TU-Berlin.DE 1136553088 N * lchvdlog lchvdlch 1136553092 M * lchvdlch morming all 1136553098 M * lchvdlch *ning* 1136553511 J * Morx ~That@dan75-3-82-227-60-181.fbx.proxad.net 1136553543 M * Morx Hi guys, 1136553611 M * Morx am at the end stage of vserver installation on a debian and after reboot I get this 1136553628 M * Morx root (hda0,0) 1136553637 M * Morx filesystem type is ext2fs, partition type 0x83 1136553698 M * Morx kernel /vmlinuz-2.6.12.4-vs2.0 root=/dev/hda1 ro 1136553710 M * Morx Error 15:file not found 1136553720 M * schellh bootloader lilo ? 1136553732 M * Morx i use grub 1136553753 M * schellh i guess you have a type in your kernel line, eg. the kernel file not found 1136553757 M * schellh typo 1136553774 M * Morx yes 1136553820 M * schellh dont you have a backup kernel ? 1136553851 M * Morx yes i've another kernel installed 1136553914 M * Morx how can i process ? 1136553946 M * schellh boot the other kernel and try to check wether /vmlinuz-2.6.12.4-vs2.0 - the line you have in menu-lst of grub is the real filename 1136554266 M * Morx right, that is 1136554292 M * Morx i'll add the right path 1136554303 M * Morx thanks :) 1136554306 M * schellh np 1136555689 Q * Morx Quit: See U All 1136555991 Q * khnz Read error: Connection reset by peer 1136557324 J * undefined ~undefined@adsl-68-93-109-94.dsl.rcsntx.swbell.net 1136557769 M * sod TheSeer: solved the problem 1136557809 M * sod it's really embarasing, the guy who set up the master server didn't add a loopback interface 1136557828 M * sod it didn't have a 127.0.0.1 localhost 1136557832 M * sod dumb one 1136557834 M * sod :) 1136557837 M * schellh lol 1136557844 M * SNy that's because there isn't a real loopback with vserver 1136557849 M * sod neah 1136557850 M * SNy unless they added it by now 1136557856 M * sod SNy: it's not that 1136557868 M * sod the master server didn't have a loopback 1136557874 M * SNy the master itself? 1136557876 M * SNy ouch 1136557879 M * sod yes:) 1136557883 M * sod ouch indeed 1136557908 M * sod noone noticed that, and that's where the weird problem emerge 1136557934 M * sod basically, you could connect to any service from a vserver from the outside but not from the vserver 1136557953 M * sod because, obvious, it tried to use the loopback 1136557967 M * sod ant that wasn't avalable 1136557985 M * sod and that's why tcpdump was so silent :)) 1136557997 M * sod I say again, really dumb 1136558022 M * SNy what is the host running? 1136558037 M * sod debian sarge 1136558037 M * SNy (which distro) 1136558039 M * SNy ah 1136558058 M * sod it isn't the distro's fault 1136558067 M * SNy well, don't say that 1136558075 M * SNy I installed a sarge within a chroot 1136558085 M * sod one of the guys that install the servers made a dumb copy&paste 1136558094 M * SNy and when I did that, the installer didn't setup a loopack, either 1136558097 M * sod in the interfaces file 1136558110 M * sod I guess 1136558122 M * sod or it might have been what you said 1136558143 M * sod anyway, nobody noticed the lack of 127.0.0.1 1136558146 M * SNy hehe 1136558160 M * SNy my mozilla wouldn't work 1136558176 M * SNy that's what made me aware there was something strange going on 1136558179 M * sod it became clear that it was a system problem when we downgraded to 1.9.x 1136558183 M * SNy totally stupid error message, too 1136558199 M * sod which we always used, and we got "invalid argument" erno 1136559478 N * Bertl_zZ Bertl 1136559482 M * Bertl morning folks! 1136559490 M * waldi good morning Bertl 1136559497 M * mnemoc moin Bertl 1136559507 M * mnemoc you are sleeping more these days :p 1136559537 M * Bertl defnitely .. 1136559977 A * micah wonders about synthetic vservers 1136560000 M * Bertl hey micah! 'synthetic vservers'? 1136560048 M * sod no synthetics; enlarge your vserver, all natural! 1136560059 M * micah Bertl: hey! yeah was just reading Patrice Levesque's email response to "Re: [Vserver] vservers and unionfs" 1136560173 M * micah there aren't many details -- it sounds like unification 1136560443 M * mnemoc i used vservers over unionfs for about a month 1136560460 M * mnemoc bad bad idea 1136560463 M * micah apparantly its for old-style configs, and there is a gui for creating vservers 1136560468 M * micah mnemoc: why bad? 1136560472 M * mnemoc crashes 1136560489 M * mnemoc twice a week at least 1136560499 M * micah ah, unionfs is not stable 1136560521 M * mnemoc on froze the context 1136560522 M * mnemoc and* 1136560577 M * mnemoc unionfs doesn't tolerate stress 1136560682 J * oliwel ~mail-at-o@host-62-245-151-178.customer.m-online.net 1136560733 M * Bertl micah: btw, 'your' util-vserver packages work fine on power5 (32bit) now going to test the 64bit version ... 1136560847 M * micah Bertl: if you are talking about the util-vserver package from deb.riseup.net, I wouldn't call that 'mine' yet -- I've done very little with that and have only made that available as a backport to sarge 1136560875 M * micah but if you are speaking of something else, when you say 'your', I'd like to know what you are referring to :) 1136560923 M * Bertl util-vserver 0.30.209-1 1136560944 M * Bertl welcome oliwel! 1136560949 M * micah Bertl: from where? 1136560972 M * Bertl well, if you tell me how to figure that? 1136560991 M * micah apt-cache policy util-vserver 1136561017 M * micah it is probably not correct to call that my package, I haven't finished and released my version yet 1136561028 M * micah so its probably the one I worked with ola to get into unstable 1136561079 M * Bertl 0.30.209-1 0 1136561079 M * Bertl 500 http://ftp.de.debian.org etch/main Packages 1136561079 M * Bertl 500 http://ftp.de.debian.org sid/main Packages 1136561124 M * micah yeah, thats the unstable package 1136561138 M * micah still needs a lot of work before I would call it mine :) 1136561195 M * Bertl that might be, but a) it works (contrary to older packages) and b) I said 'your' :) 1136561214 M * Bertl so I would say you already did good work there :) 1136561241 M * oliwel Bertl: knocknoc - happy new year ;) 1136561291 M * Bertl oliwel: and a good one 2u2! 1136561341 M * oliwel Bertl: thx 1136561364 M * oliwel Bertl: I send a mail to the list regardin vlan...had a look at ? Need urgently help on that 1136561390 M * Bertl not there yet ... 1136561397 M * Bertl but vlans work quite fine ... 1136561446 M * oliwel yes they work, but I have some probs with routing 1136561465 M * oliwel and with setting up the vlan interfaces inside the server :( 1136561516 M * Bertl you do not setup the interfaces _inside_ the guest 1136561534 M * oliwel Bertl: I ment inside the vserver config... 1136561545 M * Bertl and 'just' using eth0.10 or so instead of eth0 should work 1136561572 M * oliwel ??? 1136561587 M * oliwel instead of the vlan identifier ? 1136561599 M * oliwel so vlan3 is eth.3 ? 1136561608 M * oliwel sorry eth0.3 1136561630 M * Bertl that's how vlans are denoted on linux .. no? 1136561653 M * oliwel hmm - probably - ifconfig shows them as vlanX 1136561668 M * oliwel and gentoo configures them by itself, im not tjis deep in the topic... 1136561724 M * oliwel putting "vlan3" inside the "dev" entry brings up an error, that vserver does not suport the setup of vlans 1136561840 M * SiD3WiNDR vlans on linux are usually called ethX.Y but you can ifrename it ofcourse 1136561846 M * Bertl well, I assume some script renamed them 1136561870 M * Bertl oliwel: but if the interface exists, and you are adding an ip there, it should not matter 1136561886 M * Bertl oliwel: btw, what tool version do you use/have? 1136561922 M * oliwel ok i will try - hang on... 1136561933 M * oliwel util-vserver: 0.30.209; Jan 2 2006, 22:49:56 1136561934 M * Bertl interesting, seems Jacques is providing a third toolset ... 1136561937 M * oliwel Kernel: 2.6.14-vs2.0.1-gentoo 1136562082 M * oliwel ok you are tright - the interface is setup corectly but a waring is issued... 1136562173 M * oliwel so this seems to be solved 1136562181 M * oliwel second issue - routing :) 1136562201 M * oliwel anyone here has some good ideas on this 1136562249 M * oliwel nobody wants to talk with me :( 1136562285 M * micah Bertl: :) 1136562442 M * Bertl oliwel: be quiet! I'm answering your email :) 1136562451 M * oliwel Bertl: ok :=) 1136562463 M * Bertl oliwel: http://archives.linux-vserver.org/200311/0470.html 1136562499 M * schellh *sniff 1136562504 M * schellh i also have a problem 1136562524 M * schellh gentoo sparc - own kernel 2.6.14.4-vs2.1.0 1136562537 M * schellh 15:47:58 URL:http://ftp.debian.de/debian/pool/main/d/debootstrap/debootstrap_0.2.45-0.2_sparc.deb [46,468/46,468] -> "/var/tmp/debootstrap.uCIuw2/debootstrap.deb" [1] 1136562537 M * schellh I: Retrieving debootstrap.invalid_dists_sarge_Release 1136562537 M * schellh I: Validating debootstrap.invalid_dists_sarge_Release 1136562537 M * schellh E: Invalid Release file, no entry for main/binary-TI UltraSparc II (BlackBird)/Packages 1136562603 M * Bertl try prepending ARCH=sparc 1136562643 M * schellh cool thanks 1136562705 M * oliwel Bertl: hmm,... 1136562728 M * oliwel this is two year old - is this fixed or still a "valid" bug ? 1136562728 M * Bertl oliwel: let me guess, you expected more? 1136562746 M * Bertl oliwel: it's neither a bug, nor fixed, it is like networking is 1136562755 M * oliwel it explains some of the problems that I managed to solve with iptables / NAT 1136562980 M * oliwel so - it seems that you cdont have a howto or good practise solution on this stuff ? 1136562987 Q * shedi Quit: Leaving 1136563470 M * Bertl oliwel: let's define 'this stuff' shall we? 1136563621 M * schellh *g 1136563715 M * oliwel Bertl: *gg* 1136563731 M * Bertl oliwel: i.e. the main questions are: what is the scenario? how do we 'expect' it to work? what do we do to make it work? 1136563739 M * oliwel Bertl: checking my base system - recognized that even the host has problems with the routing 1136563750 M * oliwel Bertl: I will try 1136563763 M * Bertl oliwel: well, the problem is usual a misunderstanding of the linux networking 1136563766 M * oliwel So, I have on physical eth0 two vlan devices vlan3 and vlan4 1136563786 M * Bertl okay, so we can see that as three different devices 1136563812 M * oliwel vlan3 is the default route, the default route for vlan4 seems to be not set correctly by gentoo - I uess that I have to replace ipconfig by iproute2.... 1136563820 M * oliwel Yes - three devices 1136563821 M * Bertl (vlans happen at a very low layer in networking and are basically considered equivalent to normal interfaces) 1136563853 M * Bertl oliwel: on a 'typical' linux machine, there is only _one_ default gateway 1136563864 M * Bertl (although you can specify as many as you like :) 1136563886 M * oliwel Yes sure.... 1136563924 M * oliwel But I need one fpr the second vlan too 1136563935 M * oliwel oterhwise there is no gateway for it... 1136563955 Q * dreamist Quit: 1136563980 M * Bertl yeah, that's the scenario, two different networks, two different gateways 1136564001 M * Bertl and the solution for this is usually separate routing tables 1136564019 M * Bertl s/is usually/usually is/ 1136564022 M * oliwel yea and I assuemd that gentoo does this 1136564036 M * oliwel but it seems that this is not working :( 1136564039 M * Bertl well, it would be the first distro to do such complex things out of the box 1136564043 M * oliwel So its not a vserver problem 1136564056 M * Bertl no, definitely not :) 1136564066 M * oliwel as I can specify routes for eiterh vlan I assuemd its working... 1136564128 M * Bertl having two different tables, and selecting among them via source ip, will work as expected 1136564180 M * oliwel so I must figure out how this works on gentoo :( 1136564200 M * oliwel hmmmpf 1136564202 M * oliwel * Adding routes 1136564203 M * oliwel * default via 212.18.24.192 1136564204 M * Bertl I'd say it works as desribed (in the posting) 1136564205 M * oliwel RTNETLINK answers: Invalid argument 1136564226 M * Bertl the dev argument is missing :) 1136564253 M * oliwel Should yes, but makes an individual setup necessary - means maintenance and work :( 1136564268 M * Bertl ah, okay, well, that's something for #gentoo :) 1136564274 M * schellh well i think gentoo is so configurable 1136564358 M * Bertl FaUl: ping! 1136564390 M * Bertl schellh: could it be that you are _not_ using gentoo yourself? 1136564396 J * ichigo ~m@203.81.211.43 1136564485 M * Bertl welcome ichigo! 1136564512 M * schellh i was forced to use gentoo because the other 3 distros i trieg on my sparc e3500 wouldnt install 1136564530 M * schellh i was so dissappointed by debian sarge because their install kernel has not a single sbus driver in it 1136564543 M * schellh would need to create my own install cd.. anyways 1136564545 M * schellh i gotta go 1136564548 M * schellh see you monday 1136564583 M * oliwel Bertl: yes will check the vlan stuff on gentoo first and try to set this up properly 1136564590 M * Bertl schellh: k, cya! 1136564600 M * ichigo hello all. can someone please tell me what option to specify after ./configure so that pefix=/usr AND the initrd directory be /etc/init.d. normally i just make a symlink but now i building debian package and need to set this right. oh and hi all. i was in the middle of writing this message so could not reply 1136564644 M * Bertl ichigo: try --prefix=/ 1136564725 M * Bertl ichigo: but be aware that the debian defaults are quite different 1136564798 Q * Vudumen Ping timeout: 480 seconds 1136564865 M * ichigo anyone??? bertl?? some one must know what to place after ./configure ??? 1136564900 M * Bertl ichigo: hmm, do you ignore my answers on purpose? :) 1136564917 M * ichigo --prefix=/use is fine BUT is need the initrd directory to be /etc/init.d instead of /usr/etc/init.d 1136564947 M * ichigo Bertl. gome ne but you did not answer the question 1136564954 M * Bertl 17:24 < Bertl> ichigo: try --prefix=/ 1136564990 M * ichigo very well. see you 2 mins. 1136565033 M * lchvdlch ichigo: at least you could tell us which package are you working on 1136565045 M * Bertl I'd assume util-vserver :) 1136565047 J * Vudumen vudumen@perverz.hu 1136565067 M * ichigo i thought it would be obvios since i cam here... its util-vserver 1136565114 M * lchvdlch oh, sorry, I'm newbie here 1136565192 M * ichigo Bertl: if i use --prefix=/ then the /etc/init.d file is good BUT where will my binaries go. wont they be lying around in / 1136565211 M * Bertl did you read the overview at the end? 1136565298 M * daniel_hozac ichigo: --prefix=/usr --sysconfdir=/etc --initrddir=/etc/init.d 1136565305 M * daniel_hozac (IIRC) 1136565306 Q * Milf Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org 1136565322 M * ichigo yes but it does no t state anything concerning te binaries 1136565392 M * ichigo no --initrddir=/etc/init.d does not work. i already tried that. its also not listed in the ./configure --help 1136565714 M * daniel_hozac ah, right, it's --with-initrddir=/etc/init.d 1136565747 M * Bertl ichigo: do the pathes with --prefix=/ not suit you? 1136565752 M * ichigo well --prefix=/ does the trick. i still don't know where the binaries go but they are not in / or /usr/bin 1136565768 M * daniel_hozac /bin and /sbin. 1136565780 M * ichigo thanks bertl you rock 1136565798 M * Bertl ichigo: if you prefer /usr/bin you can add --bindir=/usr 1136565811 M * daniel_hozac +/bin ;) 1136565812 M * Bertl (similar for --sbindir= 1136565827 M * Bertl yeah, tx, with /usr/bin :) 1136565852 M * Bertl okay, off for dinner ... back shortly 1136565857 N * Bertl Bertl_oO 1136565863 M * daniel_hozac the library would still be in /lib. easier to just use --prefix=/usr if you want it installed there. 1136565872 M * daniel_hozac just look at how the spec does it. 1136565887 M * daniel_hozac --prefix=/usr is implied there. 1136565914 M * daniel_hozac as is --sysconfdir=/etc --localstatedir=/var etc. 1136566033 M * oliwel Bertl_oO: FYI: gentoo cant handle the routing - will try to fix this 1136566173 J * stefani ~stefani@superquan.apl.washington.edu 1136566671 Q * Doener Ping timeout: 480 seconds 1136566706 J * Doener doener@i5387D438.versanet.de 1136567087 N * Bertl_oO Bertl 1136567091 M * Bertl back now! 1136567097 M * Bertl oliwel: great! 1136567672 J * shedi ~siggi@inferno.lhi.is 1136567708 M * Bertl welcome shedi! 1136567713 M * Bertl ah, wb Doener! 1136567722 M * Bertl and of course welcome stefani! :) 1136567746 M * stefani hola 1136567747 M * Doener was just my 24h disconnect, but accidently i returned from university at the same time :) 1136567781 M * Bertl well, I hope it wasn't accidently, more a coincidence :) 1136567941 M * Doener uhm, well, whatever ;) 1136567979 M * Bertl Doener: did you see, we now have initial powerpc support ... 1136568014 M * Bertl (powerpc as in ppc64, not just ppc) 1136568041 M * Doener nice :D 1136568077 M * Bertl Doener: if you find some time, please have a look at the piled up delta's 1136568113 M * Doener I'll try to find some time after dinner. 1136568120 M * Bertl great! 1136568506 Q * lonewolff Read error: Connection reset by peer 1136568513 J * lonewolff ~lonewolff@host86-128-66-87.range86-128.btcentralplus.com 1136568565 J * Ryan-- ~ryan@85-210-8-139.dsl.pipex.com 1136568620 M * Bertl welcome Ryan--! 1136568632 M * Ryan-- Thank you. 1136568639 M * Ryan-- Hi all! 1136568682 M * Ryan-- Do the vserver patches work on only the particular kernel version they say they're for? 1136568703 M * Ryan-- Or for example, can you use a patch for 2.6.14.3 on 2.6.15 1136568726 M * Bertl really depends on the changes in the mainline kernels 1136568743 M * Bertl usually a 2.6.14.3 patch will apply fine on 2.6.14.5, but fail on 2.6.15 1136568770 M * Bertl so it's always good to look for updates 1136568783 M * Ryan-- ok 1136568849 M * Ryan-- the reason I ask is that the changelog for 2.6.15 lists 'critical fixes' to the forcedeth (nforce ethernet) driver so I'm dubious about using an older kernel 1136568871 M * Bertl well, then I'd go for a 2.6.15 kernel with the appropriate aptches 1136568891 M * Ryan-- what vserver patch should I use for 2.6.15? 1136568904 M * kilian Bertl: btw. is mips supposed to be supported? 1136568905 M * Ryan-- (if possible as a URL :) 1136568919 M * Bertl kilian: supposed yes, but only lightly tested 1136568930 M * kilian Bertl: ok, what's the fix to get the tools compile? 1136568933 M * kilian currently i'm at: 1136568934 M * kilian http://buildd.debian.org/fetch.php?&pkg=util-vserver&ver=0.30.209-1&arch=mips&stamp=1134436824&file=log&as=raw 1136568950 M * Bertl Ryan--: http://vserver.13thfloor.at/Experimental/ 1136568963 M * Bertl Ryan--: but if you wait a few minutes, you'll get a newer one 1136569051 M * Ryan-- ok ty 1136569073 M * Bertl kilian: http://greek0.net/vserver/syscall_shiny7-mips.2.diff applied to the syscall alternative 1136569088 J * BitterEnd ~b@ool-182e627d.dyn.optonline.net 1136569113 M * Bertl welcome BitterEnd! 1136569117 M * BitterEnd hello there 1136569129 M * Bertl kilian: I should have an updated implementation somewhere, give me a few minutes 1136569559 M * Bertl kilian: http://vserver.13thfloor.at/Experimental/delta-0.30.209-shiny8.diff 1136569627 M * kilian ok, let's see how that's going.. 1136569630 A * BitterEnd is having issues getting an FC2 image installed under FC4 1136569643 M * Bertl BitterEnd: what's the problem? 1136569664 M * BitterEnd here's what i did.... 1136569684 M * BitterEnd I followed the FC4 install guide and used the RPMS. 1136569693 M * BitterEnd went fine 1136569705 M * BitterEnd created a FC4 vserver 1136569713 M * BitterEnd a centos42 vserver 1136569734 M * BitterEnd then tried an fc2 1136569744 M * BitterEnd with yum, obviously failed b/c no yum config 1136569751 M * BitterEnd apt-get, same b/c no repos 1136569766 M * BitterEnd so, i was going to attempt to build the yum config 1136569805 M * BitterEnd so - i guess the question i have is, for the yum configurations, is that using the host box version of yum, or the new vserver... 1136569818 M * TheSeer despite the fact it should work, what do you want an FC2 vserver for if you already have FC4 working? 1136569829 M * Bertl nostalgy? 1136569833 M * BitterEnd need it for an older rpm version of postgres 1136569841 M * TheSeer hu? 1136569843 M * BitterEnd yay legacy :) 1136569872 M * Bertl BitterEnd: it depends (regarding the yum question), for the host packages, the host yum is used with the host config 1136569875 Q * ichigo Quit: Leaving 1136569888 M * BitterEnd hu? not sure what that means.....(sorry) 1136569891 M * Bertl BitterEnd: for the guest installation, the host yum is used with 'build' defaults 1136569904 M * Bertl (that's why you usually want a patche yum there) 1136569909 M * Bertl *patched 1136569933 M * Bertl and, if you configure internalized package management, then the newly installed guest yum will be used inside the guest 1136569965 M * BitterEnd ok 1136570024 M * BitterEnd i guess i need to research more...i had it somewhat installing and it was failing with some vrpm-preload error. so i whacked the box 1136570034 M * BitterEnd and am starting over 1136570122 M * BitterEnd unless someone knows of an apt repo for fc2...... 1136570211 M * oliwel Bertl: issue solved - sourced based routing (setup'd by hand) works perfectly 1136570428 M * Ryan-- Bertl: 'Ryan--: but if you wait a few minutes, you'll get a newer one' are those few minutes up yet? 1136571002 M * Bertl oliwel: excellent! 1136571012 M * Bertl Ryan--: almost 1136571031 M * Bertl Ryan--: test compile should finish in a minute 1136571195 M * BitterEnd so - forget about that, i'll find another option.... 1136571218 M * Bertl BitterEnd: what about using the rpm option? 1136571226 M * Bertl (and specifying the packets yourself) 1136571230 M * Bertl *packages 1136571244 M * BitterEnd yea, was thinking that after reading vserver help build --help 1136571272 M * BitterEnd when someone executes a uname inside a vserver....does it return the vserver kernel uname? or the host? 1136571277 M * Bertl the fc4 guest might give you an idea what packages 1136571298 M * Bertl BitterEnd: depends on the config ... but unames are virtualized 1136571305 M * BitterEnd Bertl: good idea, i'll check that out.... 1136571434 M * BitterEnd that's what i thought would happen as well, but the a uname in a centos42 vserver is returning the vs kernel.... 1136571442 M * BitterEnd by default does it not virtualize it ? 1136571445 M * BitterEnd [root@testcent42 ~]# uname -a 1136571445 M * BitterEnd Linux testcent42 2.6.14-1.1653_FC4.vs2.0.1.1 #1 Wed Dec 14 07:55:02 EST 2005 i686 athlon i386 GNU/Linux 1136571472 M * Bertl is your host called testcent42? 1136571509 M * BitterEnd yes 1136571520 M * Bertl Ryan--: http://vserver.13thfloor.at/Experimental/patch-2.6.15-vs2.0.1.2.diff and http://vserver.13thfloor.at/Experimental/patch-2.6.15-vs2.1.0.3.diff 1136571536 M * Bertl BitterEnd: and your guest? 1136571565 J * wohnzimmerschak schak@dslb-082-083-046-025.pools.arcor-ip.net 1136571571 M * Bertl welcome wohnzimmerschak! 1136571584 M * wohnzimmerschak good evening 1136571587 M * micah Bertl: do you know if util-vserver builds on hppa? 1136571593 N * wohnzimmerschak pzystorm 1136571593 M * BitterEnd Bertl: i tried both as an ssh into that vserver as well as by vserver enter 1136571604 M * micah I think it does 1136571643 M * pzystorm short question: is there a vserver-diff for 2.6.15? 1136571667 M * Bertl micah: it should 1136571681 M * Bertl pzystorm: http://vserver.13thfloor.at/Experimental/?C=M&O=A 1136571701 M * micah Bertl: yeah I can get it to build, but ola had it fail on purpose, so I'm looking at fixing that 1136571722 M * BitterEnd Bertl: my guest image was built as '-d centos42' 1136571733 M * Bertl micah: hmm, check out http://vserver.13thfloor.at/Experimental/delta-0.30.209-shiny8.diff 1136571742 M * pzystorm great stuff. this time i bookmark the page. because i didnt found that directory listing from 13thfloor.at 1136571747 M * Bertl micah: might be related to your issues ... 1136571767 M * Bertl pzystorm: yep, that's right, but it is linked on linux-vserver.org 1136571787 M * pzystorm ah ok 1136571792 M * micah Bertl: dont have issues yet, but I'll definately check it out, thanks 1136571807 M * Bertl BitterEnd: there are 3 option here: 1136571821 M * Bertl a) the virtualization is turned off, you see the host values 1136571835 M * Bertl b) the virtualization is on, but it uses the defaults (again host) 1136571858 M * Bertl c) the virtualization is on and the guest is allowed to change the values (and accidentially changes it back to the host values) 1136571883 M * Bertl BitterEnd: you can check them easily with the following tests: 1136571911 M * Bertl for c), enter the guest, type 'hostname hansi && uname -a' 1136571928 M * Bertl for b), use vuname to change it from the outside 1136571942 M * Bertl if both fails, then a) applies :) 1136572298 Q * oliwel Quit: Chatzilla 0.9.69.1 [Firefox 1.5/2005111116] 1136572407 A * BitterEnd wonders how to configure a) to on. 1136572436 M * micah Bertl: is -shiny8 the only patch to 0.30.209 that should be applied at this time, or are there others? 1136572567 M * Bertl micah: it's an update to the syscall code (supporting mips and hppa), so it's basically an update to my code there ... I'm pretty happy with the current 0.30.209 (except for the legacynet) and do not know of other patches 1136572592 M * Bertl ah, almost forgot, kudos for the mips stuff goes to Greek0! 1136572613 M * Bertl (complaints can be directed to me too :) 1136572647 M * micah Bertl: ok, so if I take the vanilla 0.30.209 and apply -shiny8 it will be enough? Or should I apply other patches as well? 1136572661 M * micah I will pass kudos on to both of you in the package :) 1136572682 M * Bertl it should be enough ... and I don't know of other patches (atm) 1136572781 M * micah ok, great 1136572974 Q * prae Quit: Execute Order 69 ! 1136573182 M * Ryan-- Bertl: thanks again 1136573194 M * Bertl Ryan--: you're welcome! enjoy! 1136573199 M * pzystorm Bertl: can you give me a small advice: which version should i take? how 'stable' is the 2.1.0.3? i cant tolerate some security leaks or so... should i better take the 2.0.1.2? 1136573222 M * Bertl pzystorm: not stable at all, it's the development branch :) 1136573260 M * pzystorm hmm ;) 1136573273 M * Bertl pzystorm: i.e. it will not burn your machine immediatly, but you might easily hit an oops if you are pushing the limits (which you should report) 1136573288 M * Bertl pzystorm: there are _no_ known bugs in either stable or devel 1136573323 M * Bertl (i.e. if I gain knowledge of any issue, it will get fixed immediately) 1136573421 M * pzystorm and what might happen when these "oops" occures?... a guest can access the host-fs?... a guest can be killed?... the host can be killed? 1136573575 M * Bertl typically kernel oops end in panics ... so they take down your server and leave a message on your console 1136573598 M * Bertl minor oops might end up with all kinds of issues (like unkillable processes and such) 1136573623 M * Bertl oopses usually do not influence security 1136573782 M * pzystorm oh ok. kernel panic wouldnt be good at all. then i have to use the stable branch. although they might not occur, i cant risk this. 1136574754 T * Bertl http://linux-vserver.org/ | latest stable 2.01, 1.2.10, 1.2.11-rc1, devel 2.1.0, exp 2.1.0.3, 2.0.1.2 | util-vserver-0.30.209 | 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 ;) 1136575338 M * eyck damn what a long topic 1136575409 M * TheSeer and nobody ever notices what he actually did change when he does that 1136575416 M * TheSeer :P 1136575477 M * Bertl keeps the concentrated attention ... :) 1136575678 M * BitterEnd Bertl: Could vprocunhide be contributing to my uname problem ? 1136575779 M * BitterEnd .. that is, unhiding /proc/version 1136576078 M * Bertl BitterEnd: no, because /proc/version is also virtualized 1136576088 J * jeeves ~Bob@c-24-11-171-10.hsd1.mi.comcast.net 1136576094 M * Bertl BitterEnd: did you do the tests I suggested? 1136576099 M * Bertl welcome jeeves! 1136576105 M * BitterEnd Bertl: I did. 1136576112 M * jeeves Hello B 1136576118 M * BitterEnd the guest was able to change the hostname 1136576238 M * micah Bertl: I'm trying to figure out a patch that Ola has applied to the util-vserver tools, in vserver-internal.h he comments out the following: 1136576241 M * micah #if defined __dietlibc__ 1136576242 M * Bertl BitterEnd: so that probably answers your question, no? 1136576244 M * micah extern long int syscall (long int __sysno, ...); 1136576246 M * micah #endif 1136576279 M * Bertl micah: first dietlibc is mandatory, and second, don't trust Ola :) 1136576287 M * micah Bertl: I can't figure out why he did that 1136576300 M * micah Bertl: hehe yeah, I know both those things :) 1136576361 M * BitterEnd yes, it tells me that the guest is somehow changing it.....past that, i have no idea honestly. 1136576384 M * BitterEnd i would really like to jail everything about a guest 1136576387 M * BitterEnd but unsure how. 1136576461 M * Bertl BitterEnd: that's part of the virtualization, i.e. by default the guest is allowed to change it 1136576478 M * Bertl BitterEnd: if you do not like that, just remove the appropriate context capability 1136576494 M * BitterEnd where can i find a list of those? 1136576612 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html (check the ccapabilities) 1136576641 M * BitterEnd flower power! 1136576664 A * Bertl .o( weedpage is nice :) 1136576819 J * dothebart ~willi@xdsl-213-196-243-201.netcologne.de 1136577094 M * Bertl welcome dothebart! 1136577243 Q * tudenbart Ping timeout: 480 seconds 1136578932 Q * ElPistoleroRoey Quit: Leaving 1136579057 M * BitterEnd is there a way to see what capabilities are on a given guest image? 1136579173 M * Bertl yep, 'cat /proc/virtual//status 1136579203 M * FaUl Bertl: pong 1136579242 M * mnemoc Bertl: did you decide something about lastnight 'discussion'? 1136579339 M * Bertl mnemoc: not yet, I delayed it for now ... 1136579344 Q * shedi Quit: Leaving 1136579354 M * mnemoc ok 1136579386 M * Bertl mnemoc: it's not a pressing issue, more a feature ... 1136579439 M * mnemoc do you have documented the vserver-syscalls 'API'? 1136579459 M * Bertl yep 1136579480 M * Bertl http://linux-vserver.org/Syscall+Switch+Info 1136579522 M * mnemoc thanks 1136579524 Q * Ryan-- Quit: Leaving 1136579571 M * Bertl the details for the commands can be found here 1136579639 M * Bertl *short delay* 1136579661 M * mnemoc *G* 1136579768 M * Bertl http://www.13thfloor.at/vserver/d_rel26/v2.1.0/split-2.6.14.4-vs2.1.0/31_2.6.14.4_cmdef.diff.hl 1136579777 M * mnemoc :D 1136579778 M * Bertl (wanted to link the latest version :) 1136579819 M * Bertl mnemoc: btw, hollows 'lib' implements all but the latest sheduler interfaces 1136579900 Q * pzystorm Quit: 1136580956 J * Ryan-- ~ryan@85-210-8-139.dsl.pipex.com 1136580968 M * Bertl wb Ryan--! 1136580971 M * Ryan-- ty 1136580977 M * Ryan-- I'm concerned I may have screwed up. :D 1136581005 M * Ryan-- I was following the Fedora Core 4 tutorial 1136581019 M * Ryan-- (I'm using Fedora Core 3 but there can't be much difference) 1136581028 M * Ryan-- When I run the "testme.sh" script 1136581030 M * Ryan-- I got 1136581043 M * Ryan-- Linux-VServer Test [V0.14] Copyright (C) 2003-2005 H.Poetzl 1136581044 M * Ryan-- utility 'vserver' could not be found. 1136581219 M * Ryan-- What does that mean? Are the vserver-utils not working or is it the kernel? 1136581336 M * Bertl that means that the "utility 'vserver' could not be found." :) 1136581355 M * Bertl i.e. the tools are _not_ in your path 1136581527 M * Ryan-- ah I was being silly :) 1136581558 M * Ryan-- did su instead of su - 1136581688 M * Bertl well, happens ... 1136581892 M * BitterEnd What's the difference between Vserver and Xen.... 1136581912 M * Bertl the level of virtualization 1136581962 M * Bertl Xen provides a special virtual architecture which allows you to run a kernel on 1136581984 M * Bertl (similar to Bochs, QEMU, UML and many other products) 1136582011 J * lilalinux_ ~plasma@dslb-084-058-195-019.pools.arcor-ip.net 1136582017 M * BitterEnd xen is more like an emulator then. 1136582034 M * Bertl Linux-VServer OTOH does the isolation and virtualization at the kernel interfaces 1136582080 M * Bertl which means is is less resource hungry and, if used properly much more performant 1136582109 M * BitterEnd cool 1136582129 M * BitterEnd but how can i virtualize (and not allow the guests to change) a 'uname -r' :) 1136582152 M * BitterEnd (by default) 1136582357 M * Bertl the responsible capability is VXC_SET_UTSNAME 1136582370 M * Bertl http://savannah.nongnu.org/cgi-bin/viewcvs/util-vserver/util-vserver/lib/ccaps-v13.c?rev=HEAD (check here for others) 1136582386 M * Bertl this is linked from the 'Flower Page' under ccaps 1136582391 Q * lilalinux Ping timeout: 480 seconds 1136582406 M * Bertl which is also the place to specify those capabilities (or to remove them) 1136582411 M * BitterEnd ok. 1136582414 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html (Flower Page) 1136582424 M * BitterEnd but, i guess i don't follow where to put that into configuration 1136582440 M * Bertl the responsible file is /etc/vservers//Ccapabilities 1136582449 M * Bertl *sorry ccapabilities 1136582471 M * BitterEnd can be .defaults ? 1136582482 M * BitterEnd or is defaults just used for building 1136582554 M * Bertl the config doesn't say anything about defaults, but you could easily write and add a script which is run after a guest creation, which adjusts your settings ... 1136582590 M * BitterEnd ok - not a big problem.... 1136582635 M * phreak`` Bertl: what are 2.1.0.* patches about ? "just" a continuation of 2.1.0 ? 1136582661 M * Bertl phreak``: they are experimental releases .. see them as -pre releases to the next 'real' release 1136582687 M * Bertl they will switch to -rc once they are considered release candidates 1136582706 M * Bertl (I will write something about the release scheme shortly) 1136582724 M * phreak`` ok. and only a last question :) then I'll leave you alone again .. will there be/is somewhere a BME for 2.6.15/vs2.0.1 ? 1136582728 M * BitterEnd do you know what the default ccapabilities are? or can I use !VXC_SET_UTSNAME in that file? 1136582766 M * Bertl phreak``: not yet, but 2.1.0 contains the patch ... 1136582797 J * prae ~benjamin@sherpadown.net 1136582821 M * Bertl BitterEnd: you use the string versions, so utsname, and yes it should support negation (but IIRC, it's ~ not !) 1136582834 M * Bertl welcome prae! 1136582838 M * phreak`` Bertl: I know of that ;) and I already tried to backport/port BME from 2,1.0 to 2.0.1 / 2.6.14 to 2.6.15 but I had no luck with both of them 1136582847 M * BitterEnd Bertl: cool, thanks. 1136582935 J * lilalinux__ ~plasma@dslb-084-058-209-054.pools.arcor-ip.net 1136582945 M * Bertl phreak``: http://vserver.13thfloor.at/Experimental/BME/delta-2.6.14.2-vs2.0.1-rc2-bme0.06.1.diff 1136583033 M * phreak`` Bertl: thats what http://dev.croup.de/proj/gentoo-vps/browser/vserver-sources/2.0-r1/4915_vs2.0-bme-0.06.1.patch?format=txt should be ;) 1136583040 M * Bertl it seems to me that we should start an opinion poll regarding the inclusion of BME into the stable branch (sooner or later) 1136583047 M * phreak`` s/should be/is/ 1136583055 P * meandtheshell 1136583077 M * prae hi bertl 1136583079 M * phreak`` Bertl: no need to run a poll on that :) (IMHO) 1136583092 M * phreak`` prae: heya frenchman :) 1136583118 M * prae phreak``: :))) 1136583123 M * prae phreak``: Hi My lord ;) 1136583321 Q * lilalinux_ Ping timeout: 480 seconds 1136583532 M * BitterEnd after i set a flag in the /etc/vservers//ccapabilities file, can i just to s vserver restart ? 1136583620 M * BitterEnd (ignore above) 1136583847 M * Bertl yep, should work, but the better approach (in general) is to stop the guest, then modify, then start again 1136583862 M * BitterEnd ok 1136583892 M * BitterEnd that flag does seem to not allow me to modify the hostname, but the uname -r is still wrong. 1136584005 M * Bertl define 'wrong' 1136584019 M * BitterEnd sorry - wrong = host's value 1136584064 M * Bertl first, I'd suggest you play around a little with the vuname utility, to change the various utsnames on the fly 1136584066 M * BitterEnd a vuname gives the correct value 1136584080 M * Bertl huh? 1136584125 M * BitterEnd wait. 1136584133 M * BitterEnd sorry, it's not. 1136584159 M * BitterEnd i think i'm confusing myself now :) 1136584174 A * BitterEnd 's a mess 1136584175 M * Bertl okay, show me what you do, what you get, and what you'd expect :) 1136584202 M * BitterEnd vserver-stat 1136584202 M * BitterEnd CTX PROC VSZ RSS userTIME sysTIME UPTIME NAME 1136584202 M * BitterEnd 0 46 85.3M 22M 0m09s30 0m06s25 13m31s65 root server 1136584202 M * BitterEnd 49155 2 5.6M 1.7M 0m00s20 0m00s00 4m45s73 testcent42 1136584218 M * BitterEnd vuname -g --xid=49155 1136584218 M * BitterEnd /etc/vservers/testcent42 Linux xasdf 2.6.14-1.1653_FC4.vs2.0.1.1 #1 Wed Dec 14 07:55:02 EST 2005 i686 (none) 1136584239 M * Bertl okay, the view in vserver-stat is the guest name (this will not change ever) 1136584253 M * Bertl it is the name you used to create the guest 1136584264 M * BitterEnd I expected to see: something more like 2.6.9-22.0.1.EL 1136584288 M * Bertl ah, well, that can be virtualized, but is not related to the hostname 1136584303 M * Bertl (which in your case is xasdf) 1136584306 M * BitterEnd yes 1136584312 M * BitterEnd i was testing the ccap flags :) 1136584346 M * Bertl again, if you look at the flowerpage, section 'uts' 1136584362 M * Bertl you see that there are a bunch of 'names' you can change 1136584371 M * Bertl (you can do that on the fly with vuname) 1136584372 M * BitterEnd ok, brining up my "unflowered" page 1136584481 M * Bertl vuname -s --xid 101 -t sysname=a -t nodename=b -t release=2.4.0 -t version=c -t machine=d -t domainname=e 1136584492 M * Bertl (assumed that guest test101 has xid 101 :) 1136584530 Q * jeeves Quit: Leaving 1136584537 M * Bertl gives: 1136584538 M * Bertl vserver test101 exec uname -a 1136584538 M * Bertl a b 2.4.0 c d GNU/Linux 1136584546 M * BitterEnd I see. 1136584559 M * Bertl .o( interesting is the GNU/Linux ... investigating) 1136584611 M * BitterEnd how can I make fill in those values automatically? that is, based on the kernel loaded when the vserver starts? 1136584641 M * mnemoc on the weedpage you con see the config dir structure 1136584648 M * Bertl BitterEnd: there is no kernel loaded when the guest starts :) 1136584667 M * mnemoc o.o 1136584675 M * Bertl BitterEnd: that's one of the differences to Xen :) 1136584697 M * BitterEnd oh 1136584700 M * BitterEnd oh 1136584727 M * BitterEnd i need to read more 1136584730 M * BitterEnd :) 1136584753 M * Bertl the Linux-Vserver Paper might be a good start :) 1136584768 A * BitterEnd nods 1136585209 A * BitterEnd sees the light and why uname -r says what it does :) 1136585367 M * Bertl ah, the GNU/Linux is produced by uname :) 1136585488 J * lilalinux_ ~plasma@dslb-084-058-200-202.pools.arcor-ip.net 1136585488 Q * lilalinux__ Read error: Connection reset by peer 1136585512 Q * Ryan-- Quit: Leaving 1136585576 J * mhelsley ~matthltc@bi01p1.co.us.ibm.com 1136585654 M * BitterEnd was FC1 a 2.4 kernel? 1136585746 M * Bertl no idea, does it matter? 1136585753 M * BitterEnd just curious. 1136585770 M * Bertl IIRC, they had both kernels in the beginning 1136585814 M * BitterEnd thanks for your help today, I actually have to start doing some work today now :) 1136585842 M * Bertl you're welcome! have fun! 1136585870 Q * BitterEnd Quit: That's Gold Jerry! 1136587042 J * Ryan-- ~ryan@85-210-8-139.dsl.pipex.com 1136587200 M * Ryan-- I'm back yet again! Anyone know how I can restrict the CPU usage of a guest to a certain percentage (or similar)? 1136587493 M * Bertl http://linux-vserver.org/Scheduler+Parameters 1136587516 M * Bertl (read it carefully, it's not easy to get on the first run) 1136587803 M * undefined bertl: what version of vserver support scheduling? (i think i remember something in the changelog) 1136587846 M * Bertl scheduling hopefully all, the hard-cpu scheduler is new in 2.x 1136587871 M * Bertl and we have a brand new per-cpu fair scheduler in devel (2.1.x) 1136587920 M * undefined bertl: sorry, to be more explicit, i've tried it in 1.9.5 (2.6.10) and 2.0 (2.6.12) but i could only get it to work for 2 (maybe 3) guests 1136587943 M * Bertl how so? 1136587956 M * undefined bertl: didn't look further into it because it's not a "must have" feature, just a precaution and nicety 1136587989 M * undefined bertl: i ran one instance of cpuburn in guest1 and two instances in guest2 1136588036 M * undefined bertl: guest1 = 50%, guest2 = 50% (where each cpuburn in guest2 got 25% total cpu according to vtop) 1136588069 M * Bertl sounds okay, well, modulo your config :) 1136588114 M * undefined but running 3 cpuburn instances in a third vserver made that vserver get more than 33% (can't remember the exact numbers) 1136588125 M * undefined bertl: yeah, i need to play with the config 1136588131 M * undefined bertl: don't have it on me right now 1136588164 M * undefined bertl: just curious if there were any known bugs in previous versions 1136588178 M * undefined bertl: regarding scheduling 1136588219 M * Bertl no, but the previous scheduler (before the 2.1.x one) was not as flexible as the new one (i.e. no idle scheduling and such) 1136588282 M * Bertl it also gave sometimes 'unexpected' results regarding SMP (because not per-cpu) 1136588305 M * Bertl the new one should be easier to predict and control 1136588338 M * Bertl (but it will take a little until tools catch up, I guess) 1136588414 M * undefined bertl: my aim was only to make scheduling "per-vserver" instead of "per-process" 1136588429 M * undefined bertl: i'm easy to please ;) 1136588456 J * shedi ~siggi@inferno.lhi.is 1136588482 M * Bertl undefined: yes, unfortunately this is a 'special' case from the scheduler perspective ... 1136588968 J * samuel ~samuel@laid.izisolution.com 1136588975 M * samuel hi 1136588992 M * Bertl welcome samuel! 1136589005 M * samuel Bertl, you are the man 1136589168 J * The_Wolfman The_Wolfma@82-47-78-71.stb.ubr05.wals.blueyonder.co.uk 1136589168 M * Bertl am I? 1136589174 M * Bertl welcome The_Wolfman! 1136589179 M * The_Wolfman ty 1136589238 Q * The_Wolfman Quit: —I-n-v-i-s-i-o-n— 2.0 Build 3515 1136589257 M * Ryan-- sorry to pester you again Bertl :) - I've just installed my first guest (debian testing). It tries (unsuccessfully) to umount stuff when I stop it and I'd like to be able to shutdown the vserver from within it. Are there programs/tweaks I should be adding to the guest? 1136589298 M * samuel brb 1136589299 Q * samuel Quit: User abort with 5 Ctrl-C's 1136589379 M * Bertl Ryan--: just remove all hardware related runlevel scripts inside the guest 1136589399 M * Ryan-- Bertl: ah thx that makes sense 1136589482 M * Bertl shutdown should work fine, without 'init' just use reboot/halt -f 1136589670 J * samuel ~samuel@laid.izisolution.com 1136589672 M * samuel hi 1136589686 M * Bertl wb samuel! 1136589694 M * samuel thanks Bertl 1136589720 M * samuel hey, It's not really vserveer-related, but anybody know how to add shadow-password support in debianÉ 1136589741 M * Bertl ahem, that should be there since ages? 1136589759 M * samuel yes, but in the vserver build it's not 1136589770 M * samuel (vserver build with the debian flavor) 1136589782 M * samuel thats weird 1136589783 M * Bertl let me check that ... 1136589820 M * Bertl my debian sarge guest has /etc/shadow :) 1136589851 M * samuel nulaid:~# useradd -m temp 1136589851 M * samuel nulaid:~# passwd temp 1136589851 M * samuel Enter new UNIX password: 1136589851 M * samuel Retype new UNIX password: 1136589851 M * samuel passwd: password updated successfully 1136589852 M * samuel nulaid:~# cat /etc/passwd | grep temp 1136589853 M * samuel temp:$1$J.WGgAQv$GW4NF1exh6xM/E7pMcblF1:1000:100::/home/temp: 1136589932 M * Bertl b:/# useradd -m temp 1136589933 M * Bertl b:/# passwd temp 1136589942 M * Bertl b:/# grep temp /etc/passwd 1136589942 M * Bertl temp:x:1000:100::/home/temp: 1136589956 M * Bertl grep temp /etc/shadow 1136589957 M * Bertl temp:$1$eSr0qOR8$HQOFCByT5uLrSOgPqOpnd0:13155:0:99999:7::: 1136589985 M * Bertl so either something got broke in your guest, or sarge changed recently 1136590006 M * samuel sarge got shadow support, i've several box running sarge and it's ok 1136590019 M * samuel maybe it's the vserver-debiantools 1136590029 M * Bertl ah well, if you use that crap ... 1136590034 M * samuel hehehee 1136590053 M * samuel that seem to work well... 1136590059 M * samuel what's the leet way? 1136590071 M * Bertl get util-vserver 0.30.209 from sid and use the same tools everybody uses 1136590085 P * stefani I'm Parting (the water) 1136590100 M * Bertl http://linux-vserver.org/alpha+util-vserver (short description) 1136590126 M * Bertl samuel: the problem with the debian tools is that they are not suited for 2.6 kernels and broken in strange ways 1136590143 M * Bertl samuel: but feel free to submit a bug report to the maintainer 1136590153 M * samuel f*ck, i'm sorry. 1136590172 M * samuel thanks a lot 1136590184 M * Bertl you're welcome! 1136590433 Q * Ryan-- Quit: Leaving 1136591181 J * Johnnie ~john@acs-24-154-53-16.zoominternet.net 1136591189 M * Bertl welcome Johnnie! 1136591921 J * Smutje_ ~Smutje@xdsl-87-78-1-148.netcologne.de 1136591940 M * Bertl wb Smutje_! 1136591981 M * Johnnie Hey. :)