1240791535 J * allquixotic ~sean@pool-151-196-238-66.balt.east.verizon.net 1240791712 M * allquixotic Hello! It would appear that my vserver guest does not have sysfs mounted (so HAL can't work), but the host does. How can I mount sysfs in the guest? Did I miss a kernel config param? I'm using recent 2.3 devel 1240791746 M * Bertl the question you should be asking yourself is, if you really want sysfs inside a guest :) 1240791778 M * Bertl but if you are sure you do, then you 'just' need to add it to the fstab (guest config) so that it gets mounted 1240791855 M * allquixotic a GNOME desktop over NX is really crippled in functionality (errors like crazy) in a guest without HAL (which requires sysfs)... I know there are security risks to exposing sysfs in there, but this is my "large and rather insecure guest" anyway ;-) the other guest is built like a fortress 1240791903 M * Bertl interesting ... why does GNOME over NX need a hardware abstraction layer deamon? 1240791976 M * allquixotic there are desktop apps that ship with Ubuntu which require HAL to function, even though they never end up using it. things like PulseAudio, the volume control app, and apparently even the desktop background require HAL :) 1240791982 M * allquixotic and the UI is unthemed also 1240792034 M * Bertl I think that is mostly a misconfiguration, btw, IMHO pulseaudio is a bad choice .. especially if your audio hardware will be remote, no? 1240792074 M * allquixotic not planning to support or need audio, but if the time comes, pulseaudio would be *especially* well suited to network transparent audio 1240792088 M * allquixotic it's more of a bad choice for a standard desktop :) 1240792145 M * allquixotic if it's a misconfiguration then it's simply because I installed the ubuntu-desktop metapackage to pull in all the extras for a GNOME desktop environment, and in so doing I grabbed a lot of things that won't matter over an NX session 1240792199 M * Bertl i.c., well, nevertheless, it might be worth to look into changing the config to avoid hal dependancies completely 1240792280 M * allquixotic perhaps :) I'm getting a permission denied when trying to mount sysfs (as root) to /sys... maybe I should quit while I'm ahead 1240792285 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240792301 J * cga ~weechat@82.84.190.35 1240792307 M * Bertl allquixotic: as I said, add it to the fstab in /etc/vservers// 1240792321 M * allquixotic oh, it has to be initialized on bootup? 1240792339 M * Bertl well, you can mount it now too, but that's the easiest way 1240792360 M * Bertl to mount it right now, you need to enter the guest's filesystem namespace and do the mount there 1240792368 M * allquixotic in /etc/fstab in the guest I have "none /sys sysfs rw,nosuid,nodev,noexec 0 0" 1240792374 M * allquixotic and I ran `mount /sys` as root in the guest 1240792376 M * Bertl that's ignore 1240792378 M * Bertl +d 1240792394 M * Bertl and you do not have the required capabilities to do the mount 1240792654 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240792711 Q * geb Quit: Quitte 1240792814 J * edlinuxgu ~kvirc@ool-4357de9d.dyn.optonline.net 1240792863 M * edlinuxgu Hello all. 1240793052 M * Bertl hey 1240793171 M * edlinuxgu I was looking for vserver-util-python. I have been using http://rpm.hozac.com/ with Cent-OS and yum. Is this something missing from the repo., or something to download separately? 1240793201 M * Bertl which version? 1240793284 M * edlinuxgu Cent 5 i386 1240793321 M * Bertl and the version of the util-vserver package? 1240793381 M * edlinuxgu [ecapriolo@laptop1 .Trash]$ yum list | grep vserver 1240793381 M * edlinuxgu * dhozac-vserver: rpm.hozac.com 1240793381 M * edlinuxgu util-vserver.i386 0.30.216-1.pre2793.el5 installed 1240793381 M * edlinuxgu util-vserver-build.i386 0.30.216-1.pre2793.el5 installed 1240793381 M * edlinuxgu util-vserver-core.i386 0.30.216-1.pre2793.el5 installed 1240793382 M * edlinuxgu util-vserver-lib.i386 0.30.216-1.pre2793.el5 installed 1240793382 M * edlinuxgu util-vserver-sysv.i386 0.30.216-1.pre2793.el5 installed 1240793384 M * edlinuxgu kernel-debuginfo.i386 2.6.22.19-vs2.3.0.34.1 dhozac-vserver 1240793384 M * edlinuxgu kernel-debuginfo.i686 2.6.22.19-vs2.3.0.34.1 dhozac-vserver 1240793386 M * edlinuxgu kernel-doc.noarch 2.6.22.19-vs2.3.0.34.1 dhozac-vserver 1240793386 M * edlinuxgu util-vserver-debuginfo.i386 0.30.216-1.pre2793.el5 dhozac-vserver 1240793388 M * edlinuxgu util-vserver-devel.i386 0.30.216-1.pre2793.el5 dhozac-vserver 1240793388 M * edlinuxgu util-vserver-legacy.i386 0.30.216-1.pre2793.el5 dhozac-vserver 1240793403 M * Bertl (please use paste.linux-vserver.org for everything longer than 3 lines) 1240793437 M * Bertl install util-vserver-python and you should be fine 1240793510 M * edlinuxgu Is that in the repository? 1240793561 M * edlinuxgu Or is that something I have to build from source? 1240793600 M * Bertl hmm, seems, daniel didn't build/package it ... better ask him why 1240793647 M * daniel_hozac that snapshot is from before the Python bindings were part of util-vserver. 1240793681 M * Bertl ah, that explains it ... any update in sight? 1240793696 M * daniel_hozac of course. 1240793700 M * daniel_hozac :-) 1240793809 M * edlinuxgu I can build it if you send me a link to the tar.gz . I was hunting around and was not exactly sure what I was looking for 1240793924 M * daniel_hozac http://people.linux-vserver.org/~dhozac/t/uv-testing/ 1240796793 M * edlinuxgu I too had to define MNT_DETACH to get it to build mount.h (you guys helped someone else with that) 1240797403 M * Bertl yeah, that's the problem with 'older' distros ... 1240798262 M * edlinuxgu Now for the fun part. The only thing I know about python is that it is senative to white space 1240801001 Q * edlinuxgu Ping timeout: 480 seconds 1240801113 J * ousado_ ~johnny@frnk-590c7071.pool.einsundeins.de 1240801541 Q * ousado Ping timeout: 480 seconds 1240802129 M * Bertl off to bed now ... have a good one everyone! cya! 1240802134 N * Bertl Bertl_zZ 1240805226 J * imcsk8 ~ichavero@189.155.144.92 1240806961 J * balbir_ ~balbir@122.172.43.141 1240808058 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1240808645 Q * sardyno Read error: Operation timed out 1240809142 J * sardyno ~me@pool-96-235-18-120.pitbpa.fios.verizon.net 1240811780 J * sharkjaw ~gab@149-240-82.oke2-bras6.adsl.tele2.no 1240812805 Q * sardyno Ping timeout: 480 seconds 1240813923 J * sardyno ~me@pool-96-235-18-120.pitbpa.fios.verizon.net 1240815171 Q * yarihm Quit: Leaving 1240816360 Q * infowolfe Read error: Operation timed out 1240816361 J * pmenier ~pme@LNeuilly-152-22-72-5.w193-251.abo.wanadoo.fr 1240816443 J * Pazzo ~ugelt@reserved-225136.rol.raiffeisen.net 1240816470 J * infowolfe ~infowolfe@c-76-105-242-186.hsd1.or.comcast.net 1240816670 J * cga ~weechat@62.196.2.6 1240817444 Q * nenolod Quit: my eyes cannot compute this misery. 1240817714 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1240818449 J * dna ~dna@186-204-103-86.dynamic.dsl.tng.de 1240819738 J * nenolod nenolod@petrie.dereferenced.org 1240819779 Q * imcsk8 Quit: This computer has gone to sleep 1240821778 Q * allquixotic Quit: Ex-Chat 1240821980 J * davidkarban ~david@193.85.217.71 1240822103 J * friendly ~friendly@ppp118-208-135-134.lns10.mel4.internode.on.net 1240823035 J * gnuk ~F404ror@pla93-3-82-240-11-251.fbx.proxad.net 1240823226 N * Bertl_zZ Bertl 1240823230 M * Bertl morning folks! 1240823497 M * pmenier morning Bertl 1240824201 Q * ktwilight__ Remote host closed the connection 1240824268 J * ktwilight__ ~ktwilight@91.178.155.120 1240824769 J * kir ~kir@swsoft-msk-nat.sw.ru 1240825048 M * Bertl off for now .. will bbl 1240825053 N * Bertl Bertl_oO 1240826224 Q * sharkjaw Ping timeout: 480 seconds 1240826435 Q * harobed Quit: Ex-Chat 1240829422 Q * kir Quit: Leaving. 1240829480 J * kir ~kir@swsoft-msk-nat.sw.ru 1240830726 Q * ktwilight__ Remote host closed the connection 1240830962 J * ktwilight ~ktwilight@91.178.155.120 1240833101 J * harobed ~harobed@pda57-1-82-231-115-1.fbx.proxad.net 1240835176 Q * ktwilight Remote host closed the connection 1240835965 J * mrfree ~mrfree@host-84-220-150-61.cust-adsl.tiscali.it 1240836396 J * esa bip@62.123.8.250 1240836771 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240836787 J * cga ~weechat@62.196.2.6 1240836862 J * esa` bip@62.123.8.75 1240836879 Q * esa Ping timeout: 480 seconds 1240837357 J * geb ~geb@AOrleans-253-1-63-102.w92-140.abo.wanadoo.fr 1240837531 M * geb hi 1240838786 Q * esa` Ping timeout: 480 seconds 1240839063 Q * friendly Quit: Leaving. 1240839791 Q * mrfree Quit: Leaving 1240840758 J * scientes ~scientes@75-165-65-163.tukw.qwest.net 1240841167 Q * FireEgl Read error: Connection reset by peer 1240841333 M * hijacker hi geb 1240841411 J * hparker ~hparker@2001:470:1f0f:32c:290:96ff:fe50:40fa 1240841952 J * FireEgl FireEgl@173-16-9-10.client.mchsi.com 1240842016 Q * CamargoBP Quit: CamargoBP 1240842214 Q * infowolfe synthon.oftc.net testlink-nucular.oftc.net 1240842558 J * edlinuxgu ~kvirc@216.223.13.111 1240842570 J * |edlinux| ~kvirc@216.223.13.111 1240843731 J * Igor2__ igor2@catv-80-99-91-123.catv.broadband.hu 1240843736 J * ktwilight ~ktwilight@91.178.155.120 1240843833 J * allquixotic ~sean@129-2-135-41.wireless.umd.edu 1240843938 M * allquixotic Hello :-) Anyone know if there is an effort going on to collect community contributed vserver appliances suited to various tasks? I know VMware has their site but you can build vservers from an image, so we should have the same capability for notoriously ugly things like mail server setups 1240844049 M * mugwump appliance ... isn't that shorthand for "easy taret" 1240844051 M * mugwump target 1240844100 M * mugwump or perhaps more fully, "out of date, poorly built, rarely patched, widely deployed and hence juicy 'sploit target" 1240844122 M * |edlinux| mug, thats what managers want :) 1240844173 M * |edlinux| allquixotic, I had a SVN repo that had a few for a while but I did not keep it up 1240844186 M * |edlinux| It would be a very nice thing to have 1240844243 M * allquixotic mugwump: If the appliance is based on a distro with active support from a security team (such as Debian-stable, or pick your own distro) then it only takes the occasional package updating to keep it secure -- but that is the responsibility of the deployer, not the person assembling the appliance 1240844282 M * allquixotic appliances which intentionally deny the user root access are really broken, I have seen a few of those on the VMware site... I wouldn't consider that an appliance, but more like crippleware 1240844302 M * |edlinux| allquixotic, are you looking to start something like this up? 1240844322 M * |edlinux| If you would I would like to get involved. 1240844327 Q * Igor2__ Quit: Leaving 1240844340 M * allquixotic |edlinux|: Possibly, yes. I have the bandwidth and the disk space for quite some appliances if the disk image contains a minimal set of files to deploy the appliance 1240844371 M * allquixotic I could dedicate maybe 250GB to serving images, and bandwidth isn't a problem 1240844416 M * |edlinux| most of the time by images are like 100-500 MB I can mirror if its a reasonable size. WE should also have some simple xml descriptor to talk about what is in the image, the author etc 1240844470 M * allquixotic We also need some sort of machinery to either detect the path to, or let the user enter the path to, /etc/vservers and /var/lib/vservers, just in case they have custom install paths 1240844491 M * daniel_hozac that's called "./configure". 1240844540 M * daniel_hozac (and /var/lib/vservers _is_ a custom install path) 1240844611 M * daniel_hozac but this sounds like a perfect fit for the post-install scripts i've been meaning to implement. 1240844628 M * daniel_hozac which would take care of the security aspects of it automatically. 1240844634 M * mugwump why not just distribute fai profile directories 1240844635 M * allquixotic daniel_hozac: I'm afraid I don't see the connection between the ./configure script run during the build of vserver or util-vserver and shipping vserver "appliances". 1240844655 M * daniel_hozac allquixotic: that was in response to "custom install paths". 1240844671 M * mugwump smaller and no maintenance burden 1240844678 M * mugwump s/no/less/ 1240844682 M * allquixotic Of course I know you can specify the directories like /etc/vservers by passing --sysconfdir=/some/other/path/vservers to ./configure, but that doesn't help with a script entering the process long after vserver has been compiled 1240844705 M * allquixotic err without the /vservers part to sysconfdir :) 1240844719 M * daniel_hozac that is the _only_ way to change the path to the configuration storage. 1240844773 M * allquixotic I have no interest in changing the path to the configuration storage - I do have an interest in detecting whether or not the current system's configuration storage directory is /etc/vservers -- or, more directly, to simply query some resource for what it _is_ rather than going on a hide-and-seek. 1240844818 M * allquixotic If we don't have that, then we'll just have to ask the user and default to /etc/vservers. 1240844826 M * daniel_hozac $__CONFDIR in util-vserver-vars. 1240844915 M * allquixotic looks like `vserver-info' provides the data in human readable format. is it safe to parse that file for the value, or would it be better to query it a different way (perhaps without needing to parse that human readable output)? 1240844924 M * allquixotic s/file/command 1240844955 M * daniel_hozac well, as i just said, $__CONFDIR in util-vserver-vars. also, you shouldn't need it. 1240845032 M * daniel_hozac especially not if you go for build profiles instead of tarballs. 1240845296 M * allquixotic Good point, actually. I could use rsync or fai. I was planning on using `vserver build` anyway, so I don't know why I thought I needed the path to /etc/vservers :) 1240845359 M * allquixotic One way would be to ship each appliance as basically a shell script that runs vserver rsync pointed at my server... the download from the website would be nice and quick, and if the dl gets interrupted, they can resume and rsync won't start completely over. 1240845399 M * daniel_hozac what i'm suggesting is a shell script that gets run at the end of building a regular guest, that configures it to do whatever. 1240845457 M * daniel_hozac that automatically takes care of security aspects, keeps you from needing any storage or bandwidth to speak of, and makes it a lot easier to audit. 1240845508 M * allquixotic That's a nice solution. I just worry that the script would get out of date when the distro gets updated to a new release and the appliance creator has moved on (and if the site gets any sort of interest, I'll have appliances up there that I didn't personally create, and won't have time to maintain.) 1240845526 M * allquixotic Will have to emphasize that each script works with a particular distro releas.e 1240845528 M * daniel_hozac the script would be distribution specific, fo course. 1240845529 M * allquixotic release* 1240845553 M * allquixotic Some users will undoubtedly try a script with $intended_version+1, though... gotta love it. 1240845557 M * |edlinux| In general it is going to be hard to keep appliance type applications up to date. In the end they are snapshots 1240845630 M * allquixotic Appliances can be thought of as either a snapshot/disk image, or as a process to be run at post-install, as daniel_hozac suggested... both have limitations, although at least if you do a clean install over debootstrap, you know you're getting the latest security patches to start. 1240845703 M * daniel_hozac and there's no inherent trust issue. 1240845720 M * allquixotic Most appliances require some degree of deployer interaction to tell it what its hostname is, set a few options, blah blah -- so it seems like that could be done in the "post-post install" ... after all the automated steps are done, at the end of the script 1240845737 M * daniel_hozac hostnames are set via vserver ... build. 1240845773 M * allquixotic that was just an example, but yes. 1240846388 J * imcsk8 ~ichavero@201.144.106.114 1240846447 M * |edlinux| This is something interesting. Since you can apply multiple templates to a guest it should be easy to create reusable ones in an rsync capably repository 1240846490 M * daniel_hozac that'd be the idea. 1240846510 M * daniel_hozac though personally i'd lean more towards HTTP. 1240846540 M * |edlinux| Yes. If we store them in SVN we can check them out, plus we have revisions. It is the same concept. 1240846594 M * allquixotic |edlinux|: I don't follow. I am a new vserver user.... what are templates? 1240846599 M * daniel_hozac e.g. vserver xyz build -m yum --interface .. --hostname .. --context .. -- -d centos5 -r http://.../centos5-postfix -r http://.../centos5-lighttpd+fastcgi -r http://.../centos5-rsyncd etc. 1240846619 M * allquixotic interesting, daniel_hozac :) 1240846621 M * |edlinux| All you can tar up a vserver and use it as a template for the next build 1240846654 M * |edlinux| You can even apply multiple templates in order 1240846685 M * allquixotic but templates are images of files, right? 1240846700 M * daniel_hozac tarballs in various formats, yes. 1240846710 M * |edlinux| They are tar files meant to be applied at the root level 1240846721 M * ghislainocfs21 allquixotic: you could use puppet to configure the guest and get reusable code and distribution abstraction 1240846723 M * allquixotic so you build your awesome templates, CentOS releases a CVE-fixing security patch on lighttpd, and....? 1240846748 M * daniel_hozac ghislainocfs21: you can never be completely distribution agnostic, no matter how hard you try. 1240846818 M * daniel_hozac IME, it's better to just accept the distribution specificity, and tweak accordingly. 1240846868 M * |edlinux| all: You use it like this vserver-build -m template --hostname myhost -t default-built.tar -t mystuff.tar -t yourstuff.tar 1240846891 M * allquixotic The natural temptation would be to go for Debian-based and Fedora-based distributions first if supporting an appliance that is supposed to work on multiple distros -- but, paths for stuff vary between even those two greatly, so it seems you'd need a separate appliance entirely for each distro. 1240846953 M * |edlinux| True but most things are more or less cross platform. There is an ok chance it will work on both. 1240847053 M * allquixotic after what daniel_hozac said earlier, I don't think having packages of files shipped from my server is a great idea. Rather, I would like to build up a rich set of infrastructure in some scripting language (Python? Ruby? or just bash?) that makes it easier to write a post-install script for doing appliancey stuff. Need "friendly" user interactivity for configuration, and it should be a bit more robust than "let's install a set of package 1240847053 M * allquixotic s". 1240847072 M * |edlinux| That is really up to the packager. If I were making an appliance I would use mostly static compilation and try not to involve anything distribution specific. 1240847073 M * allquixotic Anybody can install packages from a repository -- appliances should go the extra mile to do the sometimes atrocious configuration steps as much as possible. 1240847134 M * |edlinux| I agree on that, but consider the man hours and the size of the community. A good start would be some templates where you are force into centos 1240847146 M * daniel_hozac i don't think templates are wise. 1240847180 M * daniel_hozac it makes auditing really hard, requires you to trust the provider, and increases the amount of work required on the providers part. 1240847259 Q * dna Quit: Verlassend 1240847310 M * |edlinux| I understand but we might have two different targets in mind. An appliance is saying "I want something out of the box" like a mysql applicance. I trust the provider. If i could script it myself I dont want an appliance I want an auto build system. 1240847357 M * daniel_hozac what's your definition of a "mysql appliance"? 1240847366 M * daniel_hozac and everything can be scripted. if it can't be scripted, you're doing it wrong. 1240847425 M * allquixotic To satisfy the desire of wanting something out of the box, you can achieve this (as the provider) in two ways: 1240847449 M * allquixotic First, you can run the programs and tools needed to get the filesystem into the desired state on your local box, and ship the entire OS's filesystem state as a .tar as suggested 1240847467 M * allquixotic Second, you can simply let your end-user's computer do all the same work that you did to get the filesystem into that state. 1240847490 M * allquixotic And the latter way is better anyway because some things will inevitably need to change based on the individual end-user's needs, so you'll need a script to invoke environment-modifying tools anyway. 1240847521 M * |edlinux| I do not want to bother trying to define appliance. When people download a VMWare appliance they are usually downloading something to try out on their dev environment. They do not want a hardened app. 1240847554 M * allquixotic Basically this analogy is the difference between Gentoo and Debian, but for v12n. 1240847575 M * allquixotic On Gentoo, the user is "installing" packages when he emerges something. But he twiddles his thumbs while the system is not only installing it, but building it. 1240847579 M * allquixotic That's the "post-install script" method here. 1240847597 M * allquixotic On Debian, the user is also "installing" packages, but he does it just by downloading the actual bits. 1240847601 M * allquixotic That's the "template" method here. 1240847632 M * allquixotic The end result is the same if you factor out the time required to execute the script -- but I don't think we'll be doing very much compiling, so the wait shouldn't be tremendous. 1240847667 M * daniel_hozac not to insult your network capacity, but i think installing from the mirrors around the world might even be faster. 1240847706 M * |edlinux| I think it it is setup correctly it could be like cpan. Different authors and lists of mirrors 1240847756 M * allquixotic daniel_hozac: Absolutely, agreed. My server is in Texas, US -- and I know at least one of my close personal friends in Malaysia can't even get 20 KB/s from my server. 1240847986 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240848050 M * |edlinux| Daniel, btw I am a strong in favor of scripting as well, but imaging/packaging is the other side of the coin. 1240848161 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1240848196 M * ghislainocfs21 just a little question, is there a limit to the number of /dev/vroot/ device we can create ? 1240848259 M * ghislainocfs21 or, with the vdevmap a way to make guest use quota on their own disk securely ? 1240848328 M * daniel_hozac by default you're limited to 16 vroots, but you can change that with the max_vroot parameter. 1240848337 M * daniel_hozac you can increase it up to 255 vroots, IIRC. 1240848352 M * daniel_hozac vdevmap has nothing to do with quotas, so i'm not sure what you're asking. 1240848384 M * |edlinux| allquixotic: Why don't you create a wiki page were we can list the requirements we want? Then we can hash out the implementation. 1240848408 M * allquixotic a wikipage on linux-vserver.org? 1240848421 M * |edlinux| I can let you use my wiki if you want 1240848435 M * allquixotic sure... I don't have a CMS set up on my box yet 1240848451 M * daniel_hozac linux-vserver.org would probably be better. 1240848461 M * daniel_hozac but that's just my opinion. 1240848492 M * |edlinux| Ok by me. 1240848598 M * ghislainocfs21 daniel_hozac: if you can give all acces to a device with vdevmap then the guest could not use quota directly ? 1240848633 M * daniel_hozac ghislainocfs21: sure, and they can also overwrite the entire filesystem or cause various security issues. 1240848821 M * ghislainocfs21 daniel_hozac: this is their disk if the risk is limited to this block device then... 1240848950 M * ghislainocfs21 daniel_hozac: but for now i will just use vroot, the setting is at compile time or can be set afterwards ? 1240849087 M * ghislainocfs21 hum no vroot on the flower page so must be somewhere else :) 1240849237 Q * davidkarban Quit: Ex-Chat 1240849297 M * ghislainocfs21 oh isee, kernel boot parameter ! 1240850086 N * DoberMann[ZZZzzz] DoberMann 1240850206 M * daniel_hozac ghislainocfs21: it's not. they can overwrite any arbitrary device they feel like. 1240850217 M * daniel_hozac kernel memory, every disk on the system, etc. 1240850348 M * ghislainocfs21 daniel_hozac: oh ok i was thinking it was limited to the block device you gives it too :) 1240850403 M * ghislainocfs21 daniel_hozac: the problem is that vroot do not behave like normal disk. For quota you cannot say "repquota /" 1240850438 M * ghislainocfs21 daniel_hozac: you must say "repquota /dev/hdv1" this makes it incompatible with quiote some tools (or i miss something :) ) 1240850441 N * DoberMann DoberMann[PullA] 1240850483 N * Bertl_oO Bertl 1240850488 M * Bertl back now ... 1240850494 M * ghislainocfs21 daniel_hozac: seems the quota tools cannot make the mapping between the device and the mountpoint 1240850498 M * daniel_hozac ghislainocfs21: that's something you can fix with vdevmap's remapping. 1240850517 M * ghislainocfs21 daniel_hozac: ohhh, interesting, could you give a quick exemple ? 1240850525 M * daniel_hozac ghislainocfs21: i.e. you give the guest the real sdXY, but remap it to vrootZ. 1240850536 J * Hawq hawk@furud.net 1240850568 M * daniel_hozac vdevmap --xid --set --open --create --device /dev/sdXY --remap --target /dev/vrootZ 1240850602 M * Hawq hi. debootstrap scripts were removed from current trunk of util-vservers. installing debian guest works fine but how one should install ubuntu guests now? 1240850626 M * Hawq btw, debootstrap stuff is still referenced in Makefile.am 1240850633 J * ghislainocfs2 ~Ghislain@adsl2.aqueos.com 1240850706 Q * harobed Ping timeout: 480 seconds 1240850707 M * daniel_hozac use a debootstrap version that knows about Ubuntu. 1240850721 M * daniel_hozac like the one pointed at by the utils. 1240850806 J * ghislainocfs22 ~Ghislain@adsl2.aqueos.com 1240850825 M * ghislainocfs22 daniel_hozac: seems there is quite a lot of inredible features that we find talking here. I try to make wiki page each time i learn because this should be shared for the benefit of all :) 1240850891 M * daniel_hozac ghislainocfs22: that was the use case the device mapping was created for. 1240850980 M * Bertl sidenote: quota should not really need the device mapping to work 1240850989 M * Hawq daniel_hozac: I'm using 1.0.12 which is ubuntu aware but no luck. 1240850999 Q * ghislainocfs21 Ping timeout: 480 seconds 1240851010 M * ghislainocfs22 daniel_hozac: i missed this in the docs 1240851026 M * ghislainocfs22 is it me or does my connection breaks every 30 seconds ? 1240851071 M * daniel_hozac Bertl: yeah, but we all know the quota tools are... "special". 1240851079 M * daniel_hozac Hawq: meaning? 1240851096 Q * bonbons Quit: Leaving 1240851114 Q * ghislainocfs2 Ping timeout: 480 seconds 1240851125 M * ghislainocfs22 Bertl: they do not work with mount point but only with device if i use vroot devices 1240851160 M * ghislainocfs22 Bertl: and so most of the tools like virtualmin use the mountpoint not the device to set quotas 1240851225 Q * ghislainocfs22 Quit: Leaving. 1240851235 J * ghislainocfs2 ~Ghislain@adsl2.aqueos.com 1240851242 Q * allquixotic Remote host closed the connection 1240851340 M * Hawq daniel_hozac: I'm on PLD and debootstrap comes from our package as well as current trunk version of util-vserver. here is result when I try any ubuntu install http://paste.linux-vserver.org/12895 where [url] is wrong (http://ftp.debian.org/debian/dists/feisty/Release). but let me see if its not PLD fault :) 1240851341 M * Bertl ghislainocfs2: a wrapper around the quota tools might do the trick nevertheless 1240851405 M * daniel_hozac Hawq: looks like you're not setting the right mirror. 1240851413 M * daniel_hozac Hawq: try -m http://ftp.ubuntu.com/ubuntu 1240851506 J * Slydder1 ~chuck@91-65-50-48-dynip.superkabel.de 1240851583 M * ghislainocfs2 Bertl: yes this is what i used but as for 127.0.0.1 better to have the tools works right away so you are not surprised by edge conditions (with my limited knowledge i create a lot of edge conditions:)) 1240851731 M * Hawq daniel_hozac: thx. it worked. 1240851850 Q * kir Quit: Leaving. 1240852042 M * Bertl ghislainocfs2: 'edge conditions' == corner cases, I presume :) 1240852638 M * Hawq daniel_hozac: btw, I've added fedora 10 to distros, patch also removes references to removed debootstrap scripts, http://hawk.furud.net/files/util-vserver-distros.patch 1240853312 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1240853410 Q * imcsk8 Quit: This computer has gone to sleep 1240853878 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net 1240854722 Q * Pazzo Quit: Ex-Chat 1240854763 J * doener ~doener@i59F5A0DC.versanet.de 1240854786 Q * harobed Ping timeout: 480 seconds 1240854869 Q * doener_ Ping timeout: 480 seconds 1240855748 J * imcsk8 ~ichavero@148.229.1.11 1240856375 J * _gh_ ~gerrit@c-71-193-204-84.hsd1.or.comcast.net 1240856713 J * cga ~weechat@82.84.190.35 1240856730 J * jrdnyquist ~jrdnyquis@slayer.caro.net 1240856731 Q * jrdnyquist 1240856944 J * Slydder2 ~chuck@91-65-50-48-dynip.superkabel.de 1240857064 Q * Slydder1 Ping timeout: 480 seconds 1240857071 Q * gnuk Quit: NoFeature 1240857210 Q * balbir_ Remote host closed the connection 1240858202 J * infowolfe ~infowolfe@c-76-105-242-186.hsd1.or.comcast.net 1240859729 Q * imcsk8 Ping timeout: 480 seconds 1240859881 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240859992 J * cga ~weechat@82.84.190.35 1240860356 J * harobed ~harobed@arl57-1-82-231-110-14.fbx.proxad.net 1240860660 Q * scientes Ping timeout: 480 seconds 1240861293 Q * infowolfe Remote host closed the connection 1240861768 J * mrfrenzy ~mrfrenzy@abydos.syntaxis.se 1240861804 M * Bertl off to bed now ... have a good one everyone! 1240861808 N * Bertl Bertl_zZ 1240862004 J * infowolfe ~infowolfe@c-76-105-242-186.hsd1.or.comcast.net 1240863930 Q * doener Read error: Connection reset by peer 1240864276 J * doener ~doener@i59F5A0DC.versanet.de 1240864794 J * dna ~dna@186-204-103-86.dynamic.dsl.tng.de 1240865209 J * imcsk8 ~ichavero@nat.ti.uach.mx 1240866104 Q * bonbons Quit: Leaving 1240866135 Q * cga Quit: got a DELL??? update you BIOS with http://github.com/cga/dellbiosupdate.sh/tree/master ;) 1240866192 Q * dna Quit: Verlassend 1240866956 Q * yang Ping timeout: 480 seconds 1240867097 J * yang ~bon@static-ip-62-75-255-125.inaddr.server4you.de 1240867465 J * saulus_ ~saulus@d003166.adsl.hansenet.de 1240867476 Q * Slydder2 Quit: Leaving. 1240867874 Q * saulus Ping timeout: 480 seconds 1240867874 N * saulus_ SauLus 1240870381 Q * phedny Ping timeout: 480 seconds 1240870527 Q * er Quit: Leaving 1240870769 N * DoberMann[PullA] DoberMann[ZZZzzz] 1240871391 Q * edlinuxgu Read error: Operation timed out 1240871561 Q * |edlinux| Ping timeout: 480 seconds 1240871948 Q * imcsk8 Quit: This computer has gone to sleep 1240872726 Q * harobed Ping timeout: 480 seconds 1240873025 J * allquixotic ~sean@pool-151-196-238-66.balt.east.verizon.net 1240873437 J * imcsk8 ~ichavero@189.155.144.92 1240874854 Q * PowerKe Ping timeout: 480 seconds 1240875429 J * PowerKe ~tom@d5153A5EC.access.telenet.be