[01:18] ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds [07:15] msc (~msc@gw2.cgn3.hosteurope.de) joined #fai. [07:15] corellis (~msc@gw2.cgn3.hosteurope.de) joined #fai. [07:15] msc (~msc@gw2.cgn3.hosteurope.de) left irc: [07:50] ErKa (keryell@keryell.pck.nerim.net) joined #fai. [08:03] siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) joined #fai. [08:15] ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds [08:33] verwilst (~verwilst@router.begen1.office.netnoc.eu) joined #fai. [08:38] ErKa (keryell@dhcp1.info.enstb.org) joined #fai. [09:22] :) [09:22] back at work [09:26] MrCatEye (me@office.metaversum.com) left #fai. [09:36] MrCatEye (~me@office.metaversum.com) joined #fai. [10:18] Nabilou (~Nabilou@mail.sysun-technologies.com) joined #fai. [10:18] Hi everybody [10:34] I want my clients to use keyboard french keys, then I created a FRENCH.var file in $FAI/config/class [10:34] But I have to idea how to use it with fai-class [10:34] to create a new class [10:34] Can anybody help please? [10:36] By the way FRENCH.var contains "KEYMAP=fr-latin9-nodeadkeys" [10:43] Nabilou: MrCatEye asked the same question yesterday while you were in the channel. maybe you can read your backlog [10:44] uchimata: ok, I didn't see that thank you [10:47] siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) left irc: Quit: Leaving. [10:55] avtobiff (~avtobiff@c-68b8e555.038-20-73746f38.cust.bredbandsbolaget.se) joined #fai. [11:02] Mrfai (~lange@suenner.informatik.uni-koeln.de) joined #fai. [11:09] this kernel line should automatically reboot after install, shouldn't it? append initrd=initrd.img-2.6.26-2-486 ip=dhcp root=/dev/nfs nfsroot=/srv/fai/nfsroot boot=live FAI_FLAGS=verbose,sshd,createvt,reboot FAI_ACTION=install [11:09] avtobiff (avtobiff@c-68b8e555.038-20-73746f38.cust.bredbandsbolaget.se) left #fai. [11:09] (generated by fai-chboot) [11:14] yep [11:14] hm [12:11] ^Miaou^ (~AntoineR@81-64-4-190.rev.numericable.fr) left irc: Ping timeout: 480 seconds [12:19] Nabilou (~Nabilou@mail.sysun-technologies.com) left irc: Ping timeout: 480 seconds [12:23] BlindBunny (~BlindBunn@power.informatik.RWTH-Aachen.DE) left irc: Ping timeout: 480 seconds [12:50] Wow! A new layout on the FAI website & ;-) [13:18] Nabilou (~Nabilou@mail.sysun-technologies.com) joined #fai. [13:52] BlindBunny (~BlindBunn@power.informatik.RWTH-Aachen.DE) joined #fai. [16:23] ErKa (keryell@dhcp1.info.enstb.org) left irc: Ping timeout: 480 seconds [17:00] Nabilou (~Nabilou@mail.sysun-technologies.com) left irc: Remote host closed the connection [17:21] Mrfai (~lange@suenner.informatik.uni-koeln.de) left irc: Quit: leaving [17:52] siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) joined #fai. [18:07] allee-k (~allee-k@alleehol.mpe.mpg.de) joined #fai. [19:08] allee-k (~allee-k@alleehol.mpe.mpg.de) left irc: Ping timeout: 480 seconds [19:34] siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) left irc: Quit: Leaving. [19:43] verwilst (~verwilst@router.begen1.office.netnoc.eu) left irc: Ping timeout: 480 seconds [19:57] serge (~serge@g25-110.citenet.net) joined #fai. [20:02] hi guys! I have FAI on Deb Etch amd64 platform. I've completed quickstart steps and looks it works. Now I'm trying to learn FAI and develop and plan my own installation. I boot Intel CPU client and looks there is a problem, as FAI has amd64 kernel installed. How can I configure FAI so I would be able to choose which kernal to boot: amd64 | i386 ? It's clear I will need this functionality in future. ...Or at least where can I take ready to use i386 [20:02] kernel for FAI to adjust things quickly? [21:09] serge: FAI just uses standard tools to boot a kernel. So whatever method you use to boot your clients (PXE, CD, ...), you can choose for yourself which kernel you want to boot by adapting the boot configuration file. [21:11] you mean changing the pxelinux settings?.. Well, I'm reading syslinux doc's now how to configure boot menu... [21:15] apparently I'll need to have separate FAI catalog and repository for amd64 and i386, then it will be able to supply it as a parameter for kernel, like: nfsroot=/srv/fai386/nfsroot or ... /srv/fai64/nfsroot, etc... correct? [21:15] I mean "It will be possible to supply.." [21:34] serge: yes [21:37] serge: in this particular case you might actually get away with having only the amd64 NFS root because FAI chroots into the system to be installed (32/64 bit) most of the time and the amd64 kernel can handle 32-bit code. That's just a guess, though. I have not tested that. [21:38] ok, thanks. [21:38] serge: and you would still face the problem of FAI always auto-detecting a 64-bit architecture [21:39] serge: I think one of the scripts in /srv/fai/config/class does that so it should be not too hard to work around that auto-detection [21:40] serge: so yes, two separate NFS roots is the easier way and I use it, too [21:40] so, the most universal choice, imho, would be to use i386 FAI binaries for initial stage..? [21:40] o, well [21:41] probably it will be the easiest way [21:41] serge: hm, not sure what you mean by i386 binaries for the initial stage. [21:42] I mean i386 kernel and i386-compatible binaries in nfsroot... If im correct in my guessing... [21:44] so when the client booting and FAI starting installation it would be compatible with any of AMD or Intel CPU... [21:44] serge: all the FAI 'binaries' (commands) that I know are scripts so no portability issues there, but I cannot guarantee that that is the case for the complete FAI tool chain [21:45] ok, I probably need to learn it more. Well, I need to go home now... Thank you. [21:46] serge: I believe that you would run into trouble when booting an i386 NFS root and tried to install an amd64 system [21:46] the other way around it may work, though [21:47] ok, then what do I need to have to be able to do both kinds of installation ? [21:48] see, the system booted with the NFS root fs is not cloned onto the client, the NFS root fs just serves as a trampoline to put parition and format the client and then put some other root FS on the client's root partition [21:49] serge: your first idea [21:50] well, I need some doc reading then... But the binaries in /srv/fai/nfsroot/bin are compiled against which kernel? x64 or i386 . Does it depends on my server(amd64) arch ? [21:51] there are some scripts and binaries also... [21:51] serge: it's kind of the dumb way but it works: install FAI on an i386 an run fai-setup to create a root FS for i386, then do the same on a amd64 system and just create the root NFS into a different directory than the i386 one [21:52] ...this is exactly what I've thought about :) :) [21:53] serge: by default, the NFS-root generated by fai-setup is the same architecture as the system you run fai-setup on [21:54] ..so the executables are taken from the system. he-he... [21:54] serge: not really [21:55] serge: FAI uses debootstrap to set up a minimal Debian root file system - just not in an actual partition but in a directory [21:55] serge: and that would be /srv/fai/nfsroot by default [21:56] serge: I'm a bit repetitive on the 'by default' thing because FAI is rather flexible in all these things :) [22:21] serge: at my work, we have both 32-bit and 64-bit nfsroots [22:22] serge: it's actually quite easy to maintain, you can have e.g. /etc/fai32/* and /etc/fai64/*, then do e.g. make-fai-nfsroot -C /etc/fai64 to generate the 64-bit nfsroot [22:23] serge: otoh, i use the same kernel and nfsroot to install both lenny and hardy, and hopefully soon lucid [22:24] serge: you can specify "--arch amd64" or "--arch i386" in make-fai-nfsroot.conf FAI_DEBOOTSTRAP_OPTS [22:40] wrench- (~wrench-@gateway-1237.mvista.com) joined #fai. [22:40] anyone awake? [22:42] yes [22:42] not that it matters - the best thing is to just write your question here and wait up to 24h [22:49] i'm trying to replicate the main FAI server here at my office and i'm running into some trouble. i have copied over all the fai config dirs and the debian mirror, but at the end of the installation i get an error about not being able to ssh to the master [22:50] so it seems that i have an ssh key authentication issue [23:06] yes, that might be the problem [00:00] --- Wed Jan 20 2010