| ErKa (~keryell@104.Red-88-2-190.staticIP.rima-tde.net) left irc: Ping timeout: 480 seconds |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) joined #fai. |
| proti (~seb@APuteaux-151-1-33-221.w82-120.abo.wanadoo.fr) joined #fai. |
| 08:49 proti | morning |
| 08:50 proti | I wonder if it's possible to use metaclass (class that defines other class) in the fai config ? |
| 08:51 oz_ | should be no problem. |
| 08:51 proti | How ? |
| 08:52 oz_ | look in /srv/fai/config/class and check out the scripts there |
| 08:54 oz_ | http://www.informatik.uni-koeln.de/fai/fai-guide/ch-config.html#s-s2 |
| 08:55 oz_ | no so much Doku about it, but the scripts show how to do it |
| sepski (~sep@40.211.jostedal.no) left irc: Ping timeout: 480 seconds |
| 08:56 oz_ | I think you can use the command 'ifclassÄ |
| 08:56 oz_ | 'ifclass', even... |
| 08:56 proti | I did. Should I do : 'ifclass <metaclass> && echo 'the classes I want to define' ? |
| 08:56 oz_ | if not, roll you own |
| 08:56 proti | In a 99-something script ? |
| 08:56 oz_ | proti: yes, should work... |
| 08:57 proti | Ok, that should work. |
| 08:57 proti | There is no more sane way to do it (like a METACLASS file with all the class in it). |
| 09:00 proti | Thanks for the tip. |
| ErKa (~keryell@193.145.150.119) joined #fai. |
| 09:05 proti | I would like to report some problem with setup-storage. |
| 09:05 proti | Should I create an account ? |
| 09:05 oz_ | account for what? |
| 09:06 proti | wiki ? |
| 09:06 oz_ | ah... yes, a good idea |
| 09:07 proti | shortly, it works good at creating, but fails to remove old disk setup (especially, when there was RAID+LVM on them). |
| sepski (~sep@40.211.jostedal.no) joined #fai. |
| 09:08 oz_ | I think I observed that too |
| 09:08 oz_ | I observed that setup-harddisks fails in the first run, but succeeds in the second |
| 09:08 oz_ | I wonder... |
| 09:08 proti | The problem is that it must activate the old setup before cleaning it up. |
| 09:09 proti | You need to setup mdadm then wgchange -a y |
| 09:09 proti | and then lvremove, and vgremove before detroying the mdadm |
| 09:09 proti | RAID* |
| 09:09 oz_ | that's different to my problem. |
| 09:10 proti | It fails because it finds an olg VG with different name on the RAID setup. |
| 09:11 proti | Second is that it cannot preserve partitions on already existing setup. (I expected that). |
| Mrfai (~lange@suenner.informatik.uni-koeln.de) joined #fai. |
| 09:13 oz_ | moin thomas :) |
| 09:13 proti | oz_: What was your problem ? |
| 09:15 oz_ | under cetain circumstances (I don'T know yet) the filesystem is created, but the install process ends anyway... |
| 09:15 oz_ | if I restart the process all runs fine... |
| 09:16 proti | It ends certainly because some of the command issued by the setup-storage failed (exit code != 0) certainly. |
| 09:17 oz_ | I will continue to work on it this afternoon...I'll see then, I hope |
| 09:17 proti | Did you have the format.log of the 2 runs ? |
| 09:17 oz_ | yes, but not here |
| 09:18 oz_ | "here" = @work, "there" = my lab :) |
| 09:18 proti | ok |
| 09:19 oz_ | fai @ work is 100% debian and runs perfekt |
| 09:19 oz_ | fai @ lab is ubuntu and needs some work. ;) |
| 09:19 proti | I worked on 100% ubuntu. |
| 09:20 proti | grub is a nightmare on ubuntu |
| 09:20 oz_ | hehe...what makes you think so? |
| 09:21 proti | the update-grub was generating empty menu.lst |
| 09:21 oz_ | ah, yes |
| fai-guy_ (~fai-guy@p4FC23521.dip0.t-ipconnect.de) joined #fai. |
| 09:21 proti | so, no boot. |
| torkel (torkel@ip64.degernas.se) joined #fai. |
| 09:21 proti | I have a workaround after 2 days tries. |
| sts_ (~sts@hmm.ono.at) joined #fai. |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) got netsplit. |
| sts (~sts@hmm.ono.at) got netsplit. |
| torkel_ (torkel@ip64.degernas.se) got netsplit. |
| maswan (maswan@kennedy.acc.umu.se) got netsplit. |
| h01ger (~holger@socket.layer-acht.org) got netsplit. |
| maswan (maswan@kennedy.acc.umu.se) returned to #fai. |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) returned to #fai. |
| sts (~sts@hmm.ono.at) returned to #fai. |
| torkel_ (torkel@ip64.degernas.se) returned to #fai. |
| 09:22 oz_ | proti: what's the workaround? |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) left irc: Ping timeout: 481 seconds |
| sts (~sts@hmm.ono.at) left irc: Read error: Connection reset by peer |
| torkel_ (torkel@ip64.degernas.se) left irc: Ping timeout: 481 seconds |
| 09:23 proti | 1- Install a device.map |
| 09:23 proti | 2- in the postinst do a cat /var/run/grub/menu.lst-ucf* >> /boot/grub/menu.lst |
| 09:24 oz_ | grub/menu.lst-ucf? |
| 09:24 proti | Yes |
| 09:24 proti | in /var/run/grub after the update-grub |
| 09:25 proti | The kernel list is generated in this file. But not reported in the /boot/grub/menu.lst |
| 09:25 proti | go figure... |
| 09:26 oz_ | where does the *ucf suffix come from? |
| 09:26 proti | from ucf package :-) |
| 09:27 oz_ | and how comes that the /boot/grub/menu.lst isn't there? |
| 09:28 proti | There is already a good explanation, see bug #239674 on launchpad |
| 09:28 proti | https://bugs.launchpad.net/ubuntu/+source/grub/+bug/239674 |
| 09:32 oz_ | okay, so I now have to say thak you |
| 09:32 oz_ | I think that fixes something... |
| 09:32 oz_ | strange. |
| 09:32 proti | yes, strange |
| h01ger (~holger@socket.layer-acht.org) got lost in the net-split. |
| 09:33 proti | The cat is supposed to update manualy the menu.lst. |
| 09:33 oz_ | "Reverting to the Debian behavior, which we've never had any trouble with, would be an improvement over the current Ubuntu version for us." <- giggle |
| 09:33 oz_ | ubuntu could focus on configuring the desktop, and maybe the boot process |
| 09:34 oz_ | dunno why they do so many other things |
| 09:34 proti | This guy has a blog where he is a bit more explicit on the decision ubunutu made. |
| Action: oz_ just does an update and wonders if it fails also |
| siggg (~jean@tra78-1-87-88-161-80.dsl.club-internet.fr) joined #fai. |
| 09:34 proti | http://www.eyrie.org/~eagle/journal/2008-06/004.html |
| 09:34 oz_ | last update, had to reboot with an older(!!) kernel to get locales installed correctly |
| 09:35 proti | I fixed this too. |
| 09:35 oz_ | I wonder if this will happen now again |
| 09:35 siretart | you might consider poking slangasek with this bug |
| 09:35 oz_ | slangasek? |
| 09:35 oz_ | it's the kernel |
| 09:35 siretart | no, I was talking about the grub issue |
| 09:36 proti | to set a locale in ubuntu : fcopy /etc/environment; fcopy /etc/default/locale; locale-gen fr-FR.UTF-8 |
| 09:36 oz_ | ah, okay |
| 09:38 proti | I don't really get the way hwdetect is working |
| 09:38 proti | in class/20-hwdetect.source |
| ErKa (~keryell@193.145.150.119) left irc: Ping timeout: 480 seconds |
| 09:39 proti | it does not output any class. |
| 09:39 oz_ | it does |
| 09:39 proti | where ? |
| 09:40 oz_ | e.g.: |
| 09:40 oz_ | if discover --version | grep "discover \(version \)*1\." ; then |
| 09:40 oz_ | VIDEODRIVER=$(discover --xdriver video 2> /dev/null) |
| 09:40 oz_ | else |
| 09:40 oz_ | VIDEODRIVER=$(discover --data-path=xfree86/server/device/driver display) |
| 09:40 oz_ | fi |
| 09:40 oz_ | [ -n "$VIDEODRIVER" ] && echo "VIDEODRIVER=$VIDEODRIVER" >> $LOGDIR/addition |
| 09:40 oz_ | al.var |
| 09:40 oz_ | echo "Video card detected: $VIDEODRIVER" |
| 09:40 oz_ | ups. |
| 09:40 oz_ | that was a bad idea |
| 09:40 proti | Hum, I don't have the same. |
| 09:41 proti | Mine only modprobes the hardware and set dma on disk. |
| 09:41 oz_ | that's my debian fai |
| 09:42 proti | Comes from /usr/share/fai-doc/examples/simple/class/20-hwdetect.source |
| 09:43 proti | What is the one you're talking about ? |
| 09:44 oz_ | ? |
| 09:44 proti | The one with the discover stuff ? |
| 09:46 oz_ | as you said, I took it from the simple examples |
| ErKa (~keryell@193.145.150.119) joined #fai. |
| 09:47 proti | ok |
| 09:48 proti | It's not present on ubuntu. |
| 09:48 proti | All lines commented in a file DEMO/10-misc |
| 09:48 oz_ | neither in debian |
| 09:48 oz_ | but the hwdetect can work as all other scripts to |
| 09:49 oz_ | too, even |
| 09:51 proti | Ok, i'll figure something base on this. |
| 09:51 proti | based* |
| 09:52 oz_ | haha |
| 09:52 oz_ | "hwdb-xml -a" ? |
| 09:54 proti | nice, I didn't know this one. |
| 09:56 oz_ | some things are nice with ubuntu |
| 09:59 proti | It is not installed on nfsroot, discover yes |
| 10:00 oz_ | just add it to NFSROOT and run make-fai-nfsroot |
| 10:03 proti | Hum, it pulls quite a lot of dependencies |
| 10:04 proti | hal, dbus, consolekit, policykit, x11-libs, python-central, glib, acl, libsmbios. |
| 10:09 siretart | oz_: proti: are you installing hardy or intrepid? |
| 10:09 proti | hardy |
| 10:09 siretart | ok |
| Action: siretart needs to do the next fai merge, targeted for intrepid |
| 10:10 proti | intrepid is not in a good shape for installation at this time. |
| 10:11 proti | I think, I'll start fai install intrepid when they'll stabilize the distrib at the end of september |
| 10:12 siretart | well, feature freeze for intrepid is this thursday |
| 10:12 siretart | I upgraded my laptop yesterday, and was pleasently surprised |
| 10:13 proti | My home machine is intrepid. |
| 10:13 proti | Bad splash boot, gnome session does not reboot or halt. |
| 10:14 proti | And at the shutdown, it emits a lot of beeps, which is annoying as hell. |
| 10:14 siretart | the latter was even release-noted, right |
| 10:15 proti | Those problems went early in the developpement, and are still not fixed. |
| 10:17 proti | Who wrote setup-harddisk |
| 10:17 proti | ? |
| 10:17 proti | setup-storage * |
| 10:18 oz_ | # $Id: setup_harddisks 4240 2007-03-07 13:50:11Z lange $ |
| 10:18 proti | thanks. |
| 10:18 oz_ | thomas did |
| 10:20 proti | alone ? |
| 10:20 oz_ | ask him... :) |
| 10:20 oz_ | dunno but I think so, yes |
| fai-guy_ (~fai-guy@p4FC23521.dip0.t-ipconnect.de) left irc: Quit: Lost terminal |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) joined #fai. |
| 10:22 proti | Mrfai: setup-storage at the moment, fails to discard preexisting RAID and LVM volumes. |
| 10:22 proti | I'm afraid this issue won't be easyly solved. |
| 10:29 proti | brb (lunch) |
| 10:31 Mrfai | proti: I did not wrote setup_harddisk. Read the header of the file itself, then you see that I only made some changes to it. |
| 10:32 proti | right |
| 10:32 proti | Who maintains the setup-storage them ? |
| 10:32 proti | then* |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) left irc: Quit: Lost terminal |
| 10:35 proti | Wiping the partition table is not enough to get rid of RAID volumes neither LVM. |
| 10:35 proti | You have to get rid of the stored metadata too. |
| 10:36 proti | And to preserve some volumes, you have to activate all of them, and discard what you don't like. |
| 10:36 Mrfai | proti: michael tautschnig wrote and maitains setup-storage |
| stockholm (~andreas@marta.schuldei.org) left irc: Quit: bondage |
| 10:38 proti | which nick has he ? |
| ErKa (~keryell@193.145.150.119) left irc: Ping timeout: 480 seconds |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) joined #fai. |
| ErKa (~keryell@193.145.150.119) joined #fai. |
| 11:27 Mrfai | proti: mt, he's currently not online |
| ErKa (~keryell@193.145.150.119) left irc: Ping timeout: 480 seconds |
| ErKa (~keryell@193.145.150.119) joined #fai. |
| Mrfai (~lange@suenner.informatik.uni-koeln.de) left irc: Quit: leaving |
| h01ger (~holger@socket.layer-acht.org) joined #fai. |
| ajf__ (~alex@82-70-11-70.dsl.in-addr.zen.co.uk) joined #fai. |
| 14:56 ajf__ | hi |
| 14:56 ajf__ | I've hit an INTERNAL ERROR in setup-storage |
| 14:59 ajf__ | Table header not seen yet |
| 15:01 ajf__ | http://paste.debian.net/15626/ |
| 15:03 oz_ | are you running fai from an installserver or from a CD? |
| 15:04 ajf__ | an installserver |
| 15:04 eartoast | moin |
| 15:04 ajf__ | FAI 3.2.9 |
| 15:05 eartoast | still got a few photos from the FAI WS in Essen... anyone interested? |
| Action: oz_ is |
| 15:05 oz_ | the train was still running? |
| 15:07 ajf__ | I've tried starting with a blank partition table, but the parsing still fails. |
| 15:08 oz_ | what driver are you using for the DELL PERC 6/i |
| 15:08 oz_ | ? |
| 15:11 ajf__ | sd? |
| 15:12 eartoast | train still working, sometimes getting stuck somewhere, but ;) |
| 15:12 ajf__ | http://paste.debian.net/15628/ |
| 15:15 oz_ | ajf__: iirc this dell card still is somewhat flacky under linux... |
| 15:15 oz_ | read e.g. http://ubuntuforums.org/showthread.php?t=719556&page=3 |
| 15:15 ajf__ | seems to work fine if I don't use setup-storage though |
| 15:15 oz_ | I think you need a custon kernel, still...but I maybe wrong don't have to much time to check today |
| 15:16 oz_ | working out-of the box...I'm surprised to hear this |
| 15:17 ajf__ | I haven't looked at doing anything fancy with it yet though. |
| fai-guy (~fai-guy@p4FC23521.dip0.t-ipconnect.de) left irc: Quit: Lost terminal |
| 15:28 ajf__ | Line 140 of Volumes.pm |
| 15:28 oz_ | ajf__: I'm sorry never used that with debian... |
| 15:29 ajf__ | the hardware, or the new setup-storage script? |
| 15:40 oz_ | the hardware |
| 15:41 ajf__ | I don't think the problems that specific to the hardware |
| 15:42 ajf__ | when the script calls 'parted', parted prints an extra 'Model:' line to STDOUT. |
| 15:42 ajf__ | I think it's this that's causing the parsing to fail. |
| 15:44 oz_ | dunno, did you look up the Dell docu for this box? |
| 15:44 ajf__ | perhaps for regular hardware this line isn't printed? |
| 15:44 oz_ | they write a few line about the use under linux |
| 15:44 ajf__ | or perhaps I'm just running with an incompatible version of parted? |
| 15:44 oz_ | question is, why parted echoes another line |
| 15:44 oz_ | another question, even |
| 15:52 ajf__ | I'm getting somewhere |
| 15:53 ajf__ | it's the version of parted from Lenny that's the issue. |
| 15:57 ajf__ | At least the version in Etch has different output than the lastest stable release. http://paste.debian.net/15634/ |
| 15:58 oz_ | but we should fix setup-storage for that |
| 15:58 oz_ | ajf__: patches welcome ;) |
| 15:58 ajf__ | I've gone for a v. quick hack by adding | grep -v Model in relevant places |
| Action: oz_ once again wathes his ubuntu-fai-cd to boot |
| 15:59 ajf__ | The script gets further, but something else has failed. I'm looking into that now. |
| Action: oz_ 's fai-cd recovered from a b0rken sources.list, as it seems |
| 16:19 ajf__ | A quick hack of my config file and all seems to be working |
| siggg (~jean@tra78-1-87-88-161-80.dsl.club-internet.fr) left irc: Quit: Leaving. |
| ErKa (~keryell@193.145.150.119) left irc: Ping timeout: 480 seconds |
| ErKa (~keryell@193.145.150.119) joined #fai. |
| ErKa (~keryell@193.145.150.119) left irc: Ping timeout: 480 seconds |
| siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) joined #fai. |
| ErKa (~keryell@104.Red-88-2-190.staticIP.rima-tde.net) joined #fai. |
| teratoma (~teratoma@208.96.63.196) joined #fai. |
| 20:49 teratoma | so i accidentally upgraded my fai server to 3.2.9 |
| 20:49 teratoma | clients fail on 'mounting nfs server' |
| 20:49 teratoma | i get dumped to the busybox prompt |
| 20:50 teratoma | i have now downgraded my fai server to 3.2.8 |
| 20:50 teratoma | i no longer get that error |
| 21:02 teratoma | how do i get around this problem after i'm almost done booting ? |
| 21:02 teratoma | The authenticity of host 'moon (10.16.66.20)' can't be established. |
| 21:02 teratoma | RSA key fingerprint is ........... |
| 21:02 teratoma | Are you sure you want to continue connecting (yes/no)? |
| siggg (~jean@tra78-5-88-172-128-32.fbx.proxad.net) left irc: Ping timeout: 480 seconds |
| 21:45 teratoma | i wonder why at the end of my fai install |
| 21:46 teratoma | it tries to ssh to the client... not the server |
| 21:46 teratoma | that is weird |
| 21:46 teratoma | i would that line would be 'trying to connect to fai-server' |
| 21:46 teratoma | moon is the client |
| 23:12 teratoma | i guess that's wha ti get for running testing version of fai |
| ErKa (~keryell@104.Red-88-2-190.staticIP.rima-tde.net) left irc: Ping timeout: 480 seconds |
| --- Wed Aug 27 2008 |