[05:27] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [06:27] Nick change: \sh_away -> \sh [06:28] moin [06:42] ErKa (keryell@m6.wifi.enstb.org) joined #fai. [06:48] Mrfai (~lange@kueppers.informatik.uni-koeln.de) joined #fai. [07:07] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [08:00] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [08:08] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [08:34] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [08:42] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [08:53] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [09:01] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [09:17] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [09:25] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [09:37] mfl (~mfl@adonis.fnb.maschinenbau.tu-darmstadt.de) joined #fai. [09:42] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [09:50] MT_ (~MT@dove.informatik.tu-muenchen.de) joined #fai. [09:50] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer [09:57] Mrfai (~lange@kueppers.informatik.uni-koeln.de) left irc: Quit: leaving [09:57] Mrfai (~lange@suenner.informatik.uni-koeln.de) joined #fai. [10:03] hey guys, i am intending to install FAI on a series of optiplex 320s, these have problems with grub and also with every systemimager kernel i've seen [10:03] does FAI use grub for netbooting at all? [10:05] I recently did an fai install to my laptop, grub wasnt involved in the netbooting at all [10:05] no. [10:05] FAI is not using grub for netbooting [10:06] FAI 3.2 (just released yesterday) uses the plain Debian kernel. [10:06] can grub be configured to netboot though, out of curiousity? can i make a grub boot disk and do a pxe boot for a nic with no bootrom? [10:07] Mrfai: is that on the etch repository? I ask because i spent a week or two screwing around with systemimager and frankly while the software seemed good these Dells are AWFUL [10:07] don't ever buy an Optiplex 320 [10:08] Winkie: fai 3.2 ist not in the etch repository. But it's tested with etch. Add this line to your sources.list and you will get FAI 3.2 for etch: [10:09] deb http://www.informatik.uni-koeln.de/fai/download etch koeln [10:09] The easiest way to test if FAI will boot your machine, is to use the FAI CD. [10:09] Mrfai: sorry that's what i meant, whether it was in FAI's etch repository [10:09] and unfortunately we have no CD burners in this building [10:09] I'll will upload the new ISO image in a few seconds. [10:10] The FAI CD is using grub, but I don't think that this is a problem. [10:10] it would be for booting these machines i think, grub simply refuses to load [10:10] but it's fine, i can't burn it even if i wanted to [10:11] so i will proceed with netboot [10:14] MT_ (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [10:15] Winkie: did you look at #419766. Also grub and optiplex 320 problem, but it seems to be a xfs problem [10:15] Mrfai: which bug reference is that? link? :) [10:16] bugs.debian.org/419766 [10:17] also look at http://lists.us.dell.com/pipermail/linux-desktops/2007-March/000286.html [10:17] Mrfai: indeed pci=nomsi is the only way we can get them to boot, but grub refuses to load [10:19] still FAI looks to do exactly what we need anyway, with the added bonus that doing softupdates on lots of machines at once is slightly less painful than with systemimager :) [10:31] hmpf. [10:32] Action: oz_ got problems to boot some sata raid controller [10:41] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [10:47] The new ISO image of the FAI-CD 3.2 is available [10:50] meandtheshell (~markus@85.127.105.177) joined #fai. [10:51] ErKa (keryell@m6.wifi.enstb.org) left irc: Ping timeout: 480 seconds [10:59] Nick change: MT -> Guest1560 [10:59] Guest1560 (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer [10:59] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [11:04] interesting, i just ran fai-setup on a stock debain etch and got: [11:04] Can't call method "exists" on an undefined value at /usr/sbin/install_packages line 413, line 36. [11:04] 413 is if ($cache->exists($pack)) { # simple package name [11:05] hmm i may be missing AptPkg::Cache [11:06] and i was, libapt-pkg-perl i think needs to be added to depends [11:14] also getting this, no idea if it's an issue: WARNING: These unknown packages are removed from the installation list: live-initramfs [11:19] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [11:26] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [11:33] Mrfai: indeed i think live-initramfs might be important as when it boots it's complaining about missing /scripts/live [11:33] could you help me out? [11:36] Mrfai: I'll test your new image in some minutes [11:52] Winkie: add "deb http://www.informatik.uni-koeln.de/fai/download etch koeln" to /etc/fai/apt/sources.list [11:57] hm. anyone suggestion how to deal with a 82801GR/GH Raid controller? [11:58] Mrfai: added, will that sort out the live-initramfs issue? otherwise it seemed to be booting just fine surprisingly :) [11:59] Winkie: yep [12:00] Mrfai: excellent, it makes sense now i understand how to fix it :) [12:05] 621 seconds installtme with fai-cd 3.2 ;) [12:05] time, even [12:07] that's pretty impressive, i am going to batch install on at least 25 machines later i hope [12:07] i assume it's pretty easy to set up a user with customised .gconf directories with FAI etc? [12:08] Winkie: read the FAI guide and the man pages of fcopy and ftar. [12:09] Mrfai: i will be doing, thanks for the help so far FAI seems like a very nice solution, especially as it uses cfengine which has been recommended to me [12:23] Mrfai: well it seems to have gotten past that stage, however i can't get it to continue, these machines have a weird issue where linux can't do anything with the second sata port and sits there soft resetting it and getting internal errors for ages [12:23] is there a way to get FAI to stream logs? [12:26] Winkie: Did you read in the guide were the log files are written to? Do you already know how to log in the install client from remote via ssh? [12:27] AFAIK you can specify which devices should be probe (at least with IDE this is possible) [12:29] Mrfai: I read some information about it FTPing logs using a username, but i would have expected that to be part way or post install [12:29] although remote ssh login is interesting, i never thought of that [12:29] the problem is that it just fills the damn screen with ata errors and i have yet to find a way to force it not to probe :( [12:32] hmm it seems to be restarting part of the install, it's not on screen long enough to see [12:37] bah, hdb=noprobe, sdb=noprobe and ata2=noprobe are no help [12:40] Boot your install client with FAI_ACTION=sysinfo [12:42] ok one sec [12:44] just doing the same thing as usual, i'll leave it for a bit though and see what happens [12:45] it usually takes a minute or two before it gets past this with a standard debian install [12:45] but it does eventually boot [12:48] ok i figured a way to stop it i think, the bios has an option for it [12:50] mfl (~mfl@adonis.fnb.maschinenbau.tu-darmstadt.de) left irc: Ping timeout: 480 seconds [12:51] aha yes this seems to be doing it! [12:57] hmm 245s but errors [12:58] Mrfai: that's gotten much further this time, only grub is having an issue booting [13:04] Nick change: MT -> Guest1567 [13:04] Guest1567 (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer [13:04] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [13:38] Mrfai: if you still have a second, could you tell me how to easily get lilo installed instead of grub and add append="pci=nomsi" to the lilo config? i'm reading through the guide now but ideally need to show my boss this booting by 4pm :) [13:39] There's a class called LILO. Add the class LILO, remove the class GRUB. But it did not test lilo for a long time. [13:41] i'm not entirely sure how to do that, and I assume there's no class for grub2? :) [13:41] sorry to bug you like this [13:41] i promise i'll read the documentation thoroughly, it's a very nice product you've put together here [13:42] Winkie: have a look at class/50-host-classes. replace GRUB with LILO [13:44] aha perfect, and should i be using 'liloappend' or adding a lilo.conf? [13:56] anonick (~kvirc@brauer.math.nat.tu-bs.de) joined #fai. [13:57] I do not use lilo since several years, so you have decide by your own [13:57] anyone interested in booting from USB stick? [13:58] no problem, i added kappend to classes/FAIBASE.var if that was the right thing to do, and the lilo config shows that it should write it but we'll see [13:58] aha it worked :) [13:58] thanks so much Mrfai [13:59] Winkie: please don't forget the FAI questionnaire after you've finished your project [14:00] Mrfai: no problem at all, i was very impressed by how much you helped [14:00] just out of interest, what's the largest number of hosts anyone has reported? [14:05] more than 1000 [14:05] have a look at http://www.informatik.uni-koeln.de/fai/q-answers.html [14:05] damn that's pretty impressive :) [14:05] ah nice [14:06] Action: Mrfai also has to fill out the questionnaire :-) [14:06] Action: Mrfai is looking for testers for USB stick booting [14:08] well when i've gotten all these optiplexes working i'll be happy to help out [14:08] i've had some errors in the initial install though [14:08] paste your fai.log to paste.debian.net [14:11] you'll have to give me a minute :) [14:12] ah, it would appear the only errors in curent/error.log are because of BIOS issues on this board [14:12] no actual errors involved [14:12] perfect [14:16] indeed i am very impressed so far [14:16] just out of interest, i am beginning to understand how the class system works, is this purely based on hostname or can i define any neater method? the reason i ask is that half of these PCs have 4:3 screens and half have 16:10 and the 16:10s require a customised X config [14:17] now ideally i'd detect this by probing the monitor for it's make, but i assume this would be an example of a post install script and not a different class of install? [14:17] what? [14:17] haha [14:17] i have identical machines [14:17] mhh.. fuckin' highlights [14:17] all are optiplex 320s [14:17] however half have 1440x900 screens [14:17] and half have 1280x1024 screens [14:17] the 1440x900 ones need a custom X config to work (shitty dell monitors) [14:17] how do i do this, do i make them a seperate class of install or do i write a script to alter the xconfig post install? [14:18] Winkie: http://www.informatik.uni-koeln.de/fai/fai-guide/ch-config.html#s-s2 [14:19] Mrfai: i saw that, but I don't think 20-hwdetect is going to detect the monitor being different is it? [14:19] no. But there's a commandscalled dcc-probe or similar which may do this [14:19] ooh i was looking for something like that [14:20] it's called ddcprobe [14:20] ah thanks very much! [14:20] this channel is a beacon of helpfulness :) [14:20] it's in the package xresprobe [14:31] hi folks - the current guide (http://www.informatik.uni-koeln.de/fai/fai-guide/ch-config.html) --> chapter 7.1 still mentions "/srv/fai/config/basefiles". Since yesterday, I am with FAI 3.2 on DebianGNU/Linux sid - there's no .../basefiles around here - what am I missing? [14:32] I guess it's per option to use it and so I have to create it on demand - yes? [14:32] yep. the directory basefiles is empty by default and it does no exists in the package itself [14:32] just mkdir it and put something into it [14:33] ah - ok - I see - thank you [14:40] Action: Mrfai brb [14:40] Mrfai (~lange@suenner.informatik.uni-koeln.de) left irc: Quit: leaving [14:48] Mrfai (~lange@suenner.informatik.uni-koeln.de) joined #fai. [14:53] ErKa (keryell@keryell.pck.nerim.net) joined #fai. [14:55] How does FAI avoid that a default task is called if the corresponding hook (the hook belonging to that particular task) has already run before? [14:56] Action: meandtheshell just read about the partition.DISKLESS example (it's done there with /etc/fstab ... but what about the other hooks?) [14:59] sorry ... it's /tmp/fai/fstab I meant [15:04] you can call skipstask in your hook [15:09] ok - fine [15:09] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [15:12] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [15:15] Nick change: \sh -> \sh_away [15:50] sepski (~sep@217.17.211.51) joined #fai. [15:51] is the fai installation supposed to remove the pxelinux config file ? or must that be done manually after the installation ? [15:51] Mrfai (~lange@suenner.informatik.uni-koeln.de) left irc: Quit: leaving [16:25] oz_ (~oz@anubis.coresystems.de) left irc: Read error: Connection reset by peer [16:25] hi sepski! [16:25] (dunno, dont use pxe much..) [16:30] Mrfai (~lange@xdsl-213-196-250-246.netcologne.de) joined #fai. [16:33] sepski, task_chboot (in usr/lib/fai/subroutines-linux) tries to disable the config file using fai-chboot -vd [16:38] interesting, it didn't disable it on mine, but then again i didn't set up any log users or anything [16:48] same here, I think defining LOGUSER (and chown'ing the config files to him) was all that was needed to enable it [17:00] anonick, thanks :) [17:00] h01ger, just started looking into fai. i must say it looks awsome so far :) [17:02] :) [17:05] seeing if it's going to be viable to install debian edu ltsp servers using it :) [17:10] anonick (kvirc@brauer.math.nat.tu-bs.de) left #fai (So Long, and Thanks for All the Fish!). [17:12] MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds [17:15] Mrfai, see /msg [17:15] please :) [17:56] MT (~MT@88.217.5.7) joined #fai. [17:57] Bluemoon (~mj@www.bfr.bund.de) left irc: Ping timeout: 480 seconds [18:01] Bluemoon (~mj@www.bfr.bund.de) joined #fai. [18:04] MT (~MT@88.217.5.7) left irc: Ping timeout: 480 seconds [18:07] MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. [18:29] ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds [18:30] woffs (~Frank@brauchen.info) joined #fai. [18:52] Robinayor_ (~robin@suw.rkara.de) left irc: Remote host closed the connection [18:58] sepski (~sep@217.17.211.51) left irc: Quit: Leaving [19:39] ErKa (keryell@keryell.pck.nerim.net) joined #fai. [19:48] Robinayor_ (~robin@suw.rkara.de) joined #fai. [20:41] Mrfai (~lange@xdsl-213-196-250-246.netcologne.de) left irc: Quit: leaving [21:07] stockholm (~andreas@petrus.schuldei.org) joined #fai. [21:08] hi [21:09] hi stockholm [21:19] alleeHol (~ach@lapex-mcallee.mpe.mpg.de) left irc: Read error: Connection reset by peer [21:19] alleeHol (~ach@130.183.74.68) joined #fai. [21:23] allee_ (~ach@lapex-mcallee.mpe.mpg.de) joined #fai. [21:23] alleeHol (~ach@130.183.74.68) left irc: Read error: Connection reset by peer [21:30] is the beta still beta or is it released now? [21:32] i just try to use the köln developers apt-sources line to download the latest and greatest and now see that there is only the initrd deb [21:35] h01ger: ? [21:36] no idea about the koeln apt repo [21:36] build it from trunk :) or /tages/ [21:36] it should be in sid too [21:36] what is the etch repository? [21:36] Action: stockholm cries [21:36] yet another hoop [21:37] i do this on a xen machine and so far about everything that could go wrong did go wrong [21:41] isnt the etch repo the uni koeln one? [21:41] but in any case, there isnt a difference between the fai binary packages [21:41] afaik [21:41] +atm :) [21:46] h01ger: so the etch repository in köln contains the same packages (compiled for etch) then the sid one? [21:46] Action: stockholm consumes chocolate [21:46] stockholm: May I have some? [21:47] Barbarossa: ah, you here! [21:47] Action: stockholm tosses some chocolate bits to Barbarossa [21:47] Action: Barbarossa is virtually learning for any examination at friday :-/ [21:47] thanks [21:48] i feel generous: you can have a whole chocolate byte! [21:48] May I also have a nibble? ;) [21:48] sure :-) [21:48] cocolate nibbles are great [21:49] ..oO( How did he know, that the exam is about theoretical computer science and architectures ) [21:49] :) [21:49] Is this Little or Big Endian Chocolate? [21:49] this is swedish chocolate... definetly Little endian [21:54] Action: Barbarossa is going crazy about all this theory he'll most probalby never use in practice [21:59] woffs (Frank@brauchen.info) left #fai. [22:14] Lin_ (~igor@189.25.224.60) joined #fai. [22:15] ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds [22:19] Lin_ (~igor@189.25.224.60) left irc: Quit: Ex-Chat [23:17] meandtheshell (~markus@85.127.105.177) left irc: Quit: Leaving. [00:00] --- Thu Aug 23 2007