MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
ErKa (keryell@m9.wifi.enstb.org) joined #fai.
Nick change: \sh_away -> \sh
07:44 oz__how can it happen that in variables.log is BOOT_DEVICE=/dev/md1, but in disk_var.sh BOOT_DEVICE=/dev/sda ?
Nick change: oz__ -> oz
Nick change: oz -> Guest2745
Nick change: Guest2745 -> oz_
08:21 glanceHello.
08:22 glanceIm porting my fai-tree to new 3.2 style with unionfs and so on.
08:23 glanceTo start adapting installer for ubuntu-gutsy.
08:23 glanceim trying to find which part that should generate the boot-initramfs?
08:24 glanceoz_: are you running some plugin for software-raid or?
08:25 glanceits probably that one thats changes the BOOT_DEVICE when it runs.
08:32 oz_glance: year playing with SW_RAID
08:32 glancethen its that guy thats changes your BOOT_DEVICE
Mrfai (~lange@suenner.informatik.uni-koeln.de) joined #fai.
08:41 glanceok. the bug is that the nice "update-initramfs -k all" doesn't bite.
08:41 glanceyou needed to use the specific version to trigger it to create a initramfs.
08:41 oz_glance: ? i don't understand...?
08:42 glanceoz_: not your bug =)
08:42 glanceoz_: that bug is the one im mucking around with =)
meandtheshell (~markus@85.127.114.89) joined #fai.
08:53 glanceok. the case is:
08:55 glancein make-fai-nfsroot is update-initramfs diverted and replaced by /bin/true during the creation of the nfsroot and the divert is removed in the end and a initramfs is created. this is probably to speed up the nfs-root create-process so it does't regenerate the initramfs for each package it puts in the nfsroot.
08:56 glancethe problem with this is that when installing a ubuntu kernel package the kernel package postinst-scripts needs to create a initramfs then, or else the initramfs-tools state-system gets confused.
08:57 glancethen there is no state on the initramfs and it won't get created by the last update-initramfs run after the divert is removed.
09:13 zqadglance: i've got a patch for that
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
09:15 zqadbut i guess that you've fixed it pretty quick
09:16 glanceits a quite quick hack to fix it but, in that case im more intrested in a solution that fits upstream.
09:18 zqadi haven't got around examining it, but it seems like removing all update-initramfs works with both ubuntu and debian, so i guess that that works in the upstream. one would just need to check it further
09:18 zqadi mean, the divert entered between 3.1.8 and 3.2. before that update-initramfs was not diverted to true
09:19 zqadso it should work anyway :)
09:19 glancehmm..
09:19 zqadthe only reason i've found is that it saves time by just running update-initramfs once during make-fai-nfsroot..but i could be wrong.
09:20 glanceether try to get the flaw in the ubuntu's linux-image postinst fixed.
09:20 zqadyeah..
09:20 glanceor maybe do some nifty detection if its a buggy postinst and not divert update-initramfs in that case.
09:21 zqadthat could be a solution..but i'm curious as to why the divert was added in the first place
09:23 glancer4357 | lange | 2007-06-23 22:05:37 +0200 (Sat, 23 Jun 2007) | 2 lines
09:23 glancedivert update-initramfs so it's only called once
09:23 glancethats the commit-message
09:23 zqadi apparently suck at the svn-magic..oh, well
09:23 zqadthanks
09:24 glancezqad: to young to become a svn-jedi =)
09:24 zqad:)
09:31 Mrfaizqad: can you tell me how your patch looks like?
09:33 zqadMrfai: it simply don't divert update-initramfs anywhere..but as glance says, that might not be the correct solution
09:54 mikapMrfai: I've a basically working grml setup using live-initramfs :)
09:55 mikapMrfai: do you have any specific reason why you think fai dirinstall might be better than debian-live for creating the chroot system? (that's definitely not an offend, just because I'm curious :))
10:03 MrfaiI did not look deep into debian-live, but using fai dirinstall you can use the nice concepts of FAI classes for creating your chroot environment.
10:05 mikapright, debian-live provides building chroot as well AFAIK :) hm, one possible benefit of fai dirinstall would be deploying the created ISOs via network, right? :)
10:06 mikapMrfai: i'm trying to get fai-client fai-doc fai-server into grml so we can ship that by default, hopefully I can manage that (problem is just disk usage) - looks good now :)
10:07 mikapanything I could add to the todolist of grml so grml could become a great demo live-cd for FAI?
10:27 MrfaiI think a special version of grml could become the fai live CD
10:27 MrfaiThis must include the nfsroot, and a partitial mirror of Debian packages
10:27 mikapok, but at least as basic build environment it should be sufficient :)
10:28 Mrfainfsroot is about 330MB, the partitial mirror around 400 MB
10:28 mikapI'm working on the grml-live package which includes the fai-setup for building customized grml, as soon as I've first results I'll let you know
10:28 mikapok
10:29 Mrfaifine. Then I will build my own grml CD :-)
10:29 mikapyeah :)
sanso|work (~sanso@194.120.17.62) joined #fai.
10:37 oz_Mrfai: mikap: sounds great!
10:38 mikapoz_: the injection by Mrfai on froscon regarding use of live-initramfs was successfull so far, and due to the step-by-step help of Mrfai regarding "build grml-chroot via fai" the autobuild-setup seems to be closer than ever ;)
10:42 lazyb0ymikap: wow, happy to see you working on that :)
Action: lazyb0y is sad for not being able to come to froscon...
10:48 mikaplazyb0y: I was in your xen-talk but had to notice that it wasn't really you ;)
10:48 mikaplazyb0y: congrats to your xen-book, it's the best ressource regarding Xen I found and recommend it to everyone asking for Xen-docs :)
10:48 mikapnice talk by your collegue BTW
10:49 mikapthe best talk I attended on froscon; funny and interesting as well
10:53 lazyb0ygreat to hear :)
meandtheshell (~markus@85.127.114.89) left irc: Quit: Leaving.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Remote host closed the connection
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
12:05 Mrfaiglance, zqad will you send a bug report to ubuntu concerning the update-initramfs problem?
12:06 Mrfaisiretart: wir FAI 3.2 in gutsy reinkommen?
12:06 Mrfais/wir/wird/
sanso|work_ (~sanso@194.120.17.62) joined #fai.
sanso|work (~sanso@194.120.17.62) left irc: Remote host closed the connection
Nick change: MT -> Guest2759
Guest2759 (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
12:28 glanceMrfai: the thing is that in ubuntu the newest fai is 3.1.8.
12:28 glanceso the bug isn't introduced yet into ubuntu =)
12:29 glancethats why i started the discussion here instead,
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
12:36 Mrfaihow can we detect the the nfsroot is ubuntu and not Debian? Then I could implement a workaround in make-fai-nfsroot.
12:44 h01ger/etc/debian_version?
12:45 lazyb0yh01ger: at least on dapper, this still shows: testing/unstable
12:45 lazyb0y:(
12:46 lazyb0ymaybe by finding that they have upstart instead of init, or something like that
12:53 Mrfaidpkg -l |grep ubuntu ?
12:54 mikapnot reliable enough IMO
12:54 mikapat least double-check using uname or something like that
12:54 mikapdoesn't ubuntu have something like /etc/ubuntu_version?!
12:55 h01gerMrfai, doesnt work. some debian packages also have ubuntu in the version...
sanso|work_ (~sanso@194.120.17.62) left irc: Remote host closed the connection
sanso|work__ (~sanso@194.120.17.62) joined #fai.
13:04 Mrfailsb_release -is, wenn es denn installiert ist
13:07 Mrfaiist aber python und ich will das nicht auch noch in der nfsroot haben :-(
sanso|work_ (~sanso@194.120.17.62) joined #fai.
sanso|work__ (~sanso@194.120.17.62) left irc: Remote host closed the connection
sanso|work_ (sanso@194.120.17.62) left #fai (brb (probably)).
13:55 lazyb0yhmm, in /etc/reportbug.conf there is a line:"bts ubuntu"
14:09 h01gerhehe
14:17 lazyb0yhmm, und kann man nicht einfach in fai.conf sowas DIST= definieren, default ist debian?
14:21 Mrfaida es nur ein workaround sein soll, wuerde ich lieber eine neue option fuer make-fai-nfsroot einfuehren
14:22 h01gerfind ich auch besser. ist vorallem auch zukunftskompatibler wenns mal ne neue distri gibt
14:24 Mrfaih01ger: nicht so viel schreiben, erhol dich lieber. Viel Trinken (kein Bier) ;-)
14:26 h01gertee und osaft stehn im bett. und gestern+heute nichtmal kaffee getrunken..
14:45 lazyb0ytess und osaft im bett is aber bisserl eklig!
14:45 lazyb0ytee
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
Nick change: \sh -> \sh_away
Mrfai (~lange@suenner.informatik.uni-koeln.de) left irc: Quit: leaving
ErKa (keryell@m9.wifi.enstb.org) left irc: Ping timeout: 480 seconds
baldy (baldy@brain.servercrew.de) left irc: Remote host closed the connection
baldy (baldy@brain.servercrew.de) joined #fai.
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
ErKa (keryell@keryell.pck.nerim.net) joined #fai.
16:00 glancehmm...
16:00 glancegutsy + fai 3.2 == unionfs kernel nullpointer crash
ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds
MT (~MT@dove.informatik.tu-muenchen.de) joined #fai.
20:20 siretartglance: :(
20:21 siretartthat's still on my todo list :/
allee (~ach@allee.mpe.mpg.de) left irc: Quit: Konversation terminated!
allee (~ach@allee.mpe.mpg.de) joined #fai.
ErKa (keryell@keryell.pck.nerim.net) joined #fai.
ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds
MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds
--- Thu Sep 6 2007

Generated by irclog2html.pl 2.1_h01ger1 by Jeff Waugh - find it at freshmeat.net! Customized by Holger 'h01ger' Levsen