| alexanderwz (~alexander@karuna.med.harvard.edu) left irc: Ping timeout: 480 seconds |
| alexanderwz (~alexander@karuna.med.harvard.edu) joined #fai. |
| MT (~MT@ppp-82-135-67-194.dynamic.mnet-online.de) joined #fai. |
| MT (~MT@ppp-82-135-67-194.dynamic.mnet-online.de) left irc: Ping timeout: 480 seconds |
| 06:34 sobersabre | Boker tov. |
| 06:34 sobersabre | Lab vakar |
| 06:34 sobersabre | Dobriy Den |
| 06:34 sobersabre | dzen dobre |
| 06:35 sobersabre | bom dia |
| 06:35 sobersabre | Bon giorno |
| 06:35 sobersabre | Guten Morgen |
| 06:36 Robinayor_ | sobersabre: hi. |
| MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. |
| ErKa (keryell@m5.wifi.enstb.org) joined #fai. |
| Mrfai (~lange@kueppers.informatik.uni-koeln.de) joined #fai. |
| sep (~sep@89.130.6.136) joined #fai. |
| 07:40 stockholm | Mrfai: did you get my glorious patches? |
| allee (~ach@dialin-212-144-129-225.pools.arcor-ip.net) joined #fai. |
| 07:53 sobersabre | will be back in hour. |
| sobersabre (~bilbo@89-138-78-195.bb.netvision.net.il) left irc: Quit: Leaving |
| 07:57 Mrfai | stockholm: the ones you've send to the BTS? Yes. |
| 07:58 stockholm | Mrfai: it occured to me that i might not create all the path elements necessary. i think i only tried it with one new path element |
| 08:00 oz_ | moin |
| 08:11 Mrfai | stockholm: yep, I noticed that. IMO there's a perl module that can do this work, or I will do it in a small loop. |
| kugg (~jerkeby@89.130.6.136) joined #fai. |
| 08:15 stockholm | Mrfai: right |
| 08:17 h01ger | moin moin |
| 08:17 kugg | m000oooOOin |
| 08:20 h01ger | Mrfai, do you want to maintain fai in debian-edu (etch) - which basically means uploading the etch target fai release to our archive (i would need to add your key to our keyring and you would need to upload to another archive) - or would you prefer if you're out of it and i upload it? |
| 08:20 h01ger | we want it in our archive, so that having the debian-edu-archive-keyring is enough to install fai. without pinning from testing or adding another archive key.. |
| 08:23 Mrfai | h01ger: you can do this. But if it's really only a second upload for me *using a different dupload.conf) I can also do it. |
| 08:23 Mrfai | h01ger: BTW, do debian-edu has a BTS on its own? |
| 08:23 Mrfai | s/do/does/ |
| 08:26 h01ger | Mrfai, yes, we have our own bts (sigh, not happy with it, historic reasons). bugs.skolelinux.no - but i can make sure to forward everything fai related to the debian bts |
| 08:27 h01ger | Mrfai, it is only a 2nd upload with a different dupload.conf |
| 08:27 h01ger | i would need to get approval to add your key to our keyring, but this should be easy |
| 08:27 eartoast | moin |
| 08:28 eartoast | 18:46 < Mrfai> eartoast: does your fai-classes-config also detects classes used in script after the command ifclass? |
| 08:28 eartoast | Mrfai: yes |
| 08:30 Mrfai | eartoast: fine |
| 08:30 Mrfai | h01ger: ok, than I can do it. Can you send me the dupload.conf I need for uploadfing to edu? |
| 08:35 eartoast | Mrfai: there is not yet a cfengine parser... and none for perl scripts, as there is nothing as easily parsable as 'ifclass' in perl scripts... |
| 08:36 h01ger | Mrfai, yes, will do |
| 08:38 stockholm | Mrfai: and obviously -p implies -a. if the path does not exist the file does not exist for sure |
| 08:38 Mrfai | eartoast: parsing cfengine should be easy: [a-zA-Z_-0-9]+::, examples: suenner|richmodis::, CS_KOELN:: |
| 08:44 h01ger | right. a short grep in my config space revealed to "problems": |
| 08:44 h01ger | ./LAPTOP/40-apt: any:: |
| 08:44 h01ger | ./LAPTOP/40-apt: DeleteLinesStarting 'Acquire::http::Proxy ' |
| 08:45 h01ger | "any::" is not really a problem though :) |
| 08:59 stockholm | MT: ping |
| 09:00 stockholm | MT, Mrfai: es scheint so zu sein dass der trend im Kernel mehr und mehr zur "asynchronous device detection" geht |
| 09:04 MT | aber haben sie auch Lösungen für unser Problem? |
| 09:04 stockholm | nein, und sie sehen es auch nicht als echtes problem an |
| 09:05 stockholm | 10:45 <hch> people caring about topology but not uuids is a rather rare special case |
| 09:08 stockholm | das bedeutet dass wir probleme bekommen beim repartitionieren |
| 09:09 stockholm | das installieren ist relativ unkritisch |
| 09:10 stockholm | comments? |
| 09:14 MT | also ich sehe nur die Lösung mit den Seriennummern |
| 09:27 stockholm | MT: dann müssen wir irgendwo eine liste mit uuids->partitionsmapping ablegen und beim weiderholten booten finden und verwenden? |
| 09:34 MT | hmm, wie sollte das gehen? |
| 09:38 h01ger | so wie jetzt mit disk_var.sh |
| 09:42 eartoast | schaut mal an, was psibox dazu gebastelt hat... |
| 09:43 MT | wo findet man das, eartoast? |
| 09:43 eartoast | BTS |
| 09:43 eartoast | wishlist gegen fai |
| 09:44 eartoast | patches gegen setup_harddisks |
| 09:44 eartoast | unsere fstabs mounten mittlerweile alle hdd-partitionen per uuid |
| 09:45 MT | eartoast, welche bug nummer? |
| 09:45 eartoast | wattema |
| 09:45 eartoast | muss selbst erst wieder nachschauen |
| 09:46 eartoast | Bug#416633 |
| allee (~ach@dialin-212-144-129-225.pools.arcor-ip.net) left irc: Read error: Connection reset by peer |
| 09:47 eartoast | und ansonsten zu disk-id: http://www.mail-archive.com/linux-fai-devel@uni-koeln.de/msg00023.html |
| 09:47 MT | ah, eartoast |
| 09:47 MT | mounten per UUID ist ok |
| 09:48 MT | das geht mit setup_harddisks2 nativ |
| 09:48 MT | aber das paritionieren |
| 09:48 stockholm | wie geht das bei reboot vom netz? |
| 09:48 MT | bzw. bei preserve o.ä. |
| 09:48 stockholm | genau |
| 09:48 MT | das - was? |
| 09:48 stockholm | das lesen von uuids |
| allee (~ach@dialin-145-254-253-190.pools.arcor-ip.net) joined #fai. |
| 09:49 stockholm | das lesen des uuid->partitions mapping |
| 09:49 eartoast | uuid is unique |
| 09:49 eartoast | deswegen ist preserve-by-uuid etwas schwierig |
| 09:49 eartoast | man koennte / mounten |
| 09:50 eartoast | dann die fstab rauskopieren |
| 09:50 eartoast | und dann wieder unmounten |
| 09:50 eartoast | dann hast du das uuid-mountpoint-mapping |
| 09:50 stockholm | eartoast: wie weisst du wo / ist? |
| 09:50 eartoast | preserve-by-label ist IMHO besser... |
| 09:51 eartoast | stockholm: einfach jede partition einmal irgendwohin mounten |
| 09:51 eartoast | und schauen, auf welcher etc/fstab liegt |
| 09:51 eartoast | ist nicht ganz idiotensicher |
| 09:51 eartoast | aber klappt meistens ganz gut |
| 09:52 eartoast | ich benutze so einen hack, um ssh-hostkeys ueber einen reinstall herueberzuretten |
| 09:52 stockholm | man könnte auch versuchen die /bootpartion zu finden und die kernel mount options zu finden |
| 09:52 stockholm | enthält das nicht die root uuid? |
| 09:52 eartoast | stockholm: das laeuft IMHO auf die gleichen probleme hinaus |
| 09:53 stockholm | warum? |
| 09:53 stockholm | die maschiene muss ja wohl irgendwie zuverlässig /boot finden |
| 09:53 eartoast | weil genauso zufaellig auf irgendeiner anderen partition "boot/" existieren kann |
| 09:53 stockholm | wie macht sie das? |
| 09:53 eartoast | und: |
| 09:54 stockholm | boot flag für die partion könnte ein hint sein |
| 09:54 eartoast | du musst dich dann mit dem parsen einer grub-config herumschlage |
| 09:54 eartoast | n |
| 09:54 eartoast | oder lilo |
| 09:54 stockholm | haben wir ein bootflag? |
| 09:54 eartoast | fstab parst sich schoener |
| 09:54 h01ger | oder grub2 oder yaboot oder eilo oder was es noch fuer bootloader gibt |
| 09:55 h01ger | wir sind !i386 :) |
| 09:55 eartoast | bootflag braucht man nicht mehr, wenn man lilo/grub im MBR hat |
| 09:55 stockholm | den MBR lesen... wie findet man die disk mit dem mbr? |
| 09:55 eartoast | /etc/fstab ist simpler |
| 09:56 eartoast | du weisst doch: KISS |
| 09:56 stockholm | eartoast: das kann simpeler sein aber nicht eindeutig |
| 09:57 eartoast | weisst du bei zwei platten, von welcher das bios gerade bootet? |
| 09:57 eartoast | also welcher mbr von beiden der authoritative ist ? |
| 09:57 stockholm | das ist was ich gerade gefragt habe. wie findet die maschine den MBR? |
| 09:58 eartoast | erste 512b der festplatte |
| 09:58 stockholm | welcher festplatte? |
| 09:58 eartoast | jeder platte ;) |
| 09:58 eartoast | dann ist es sache des bios-setups |
| 09:58 stockholm | wir werden wohl systeme mit 24 oder 42 festplatte haben |
| 09:58 eartoast | welche platte genommen wird... |
| 09:58 Mrfai | soviel ich uuid verstehe ist das ein feature des file systems. uuid wird auch nicht von allen file systemen unterstuetzt. Somit kann uuid nicht fuer eine leere partition genutzt werden IMO. |
| 09:59 Mrfai | oder verwechsele ich das jetzt uuid und label? |
| 09:59 eartoast | Mrfai: imho hast du das prob bei beiden |
| 09:59 stockholm | ich weiss es nicht |
| 10:00 stockholm | 10:47 <hch> the raid superblock has a uuid for the raid and an identifier which of the disks this is |
| 10:00 stockholm | 10:48 <hch> but in general I'm more a fan of wwns when available then the uuids in the filesystem |
| 10:00 stockholm | keine ahnung was WWNS ist |
| 10:00 stockholm | eartoast: wenn das bios weiss welche platte die erste ist, wie macht es das? |
| 10:02 eartoast | stockholm: das hat der user so konfiguriert... indem er die bootreihenfolge festlegt |
| 10:02 stockholm | eartoast: das ist keine brauchbare antwort :) |
| 10:10 Mrfai | kleiner test: partition auf USB stick erzeugt, ohne file system drin. /proc/partition zeigt sdc1 sauber an, /dev/disk/by-uuid nicht, /dev/disk/by-id ja. Alsi uuid gibt es nur wenn ein file system in der partition drin ist. |
| Nick change: MT -> Guest2144 |
| Guest2144 (~MT@dove.informatik.tu-muenchen.de) left irc: Read error: Connection reset by peer |
| MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. |
| 10:13 h01ger | fuer welche filesysteme gibts denn uuid? wenn reiser oder xfs das nicht hat, faende ich das nicht schlimm, andererseits, vfat/ntfs ist muss.. |
| 10:14 eartoast | stockholm: es ist recht schwer, diese info aus dem cmos zu extrahieren ;) |
| 10:14 stockholm | vielleicht müssen wir anders denken was den preserv und resize prozess angeht. |
| 10:15 stockholm | wenn wir das partitionieren nur beim installieren machen und das preserve und resize und umpartitionieren mit softupdates? |
| 10:15 stockholm | eartoast: was denkst du dazu? |
| 10:16 stockholm | h01ger: wir möchten gerne uuids für XFS... |
| kugg (~jerkeby@89.130.6.136) left irc: Read error: Connection reset by peer |
| 10:26 Barbarossa | stockholm: Wofür will man UUIDs? |
| 10:30 stockholm | um harddisks zuverlässig wiederzufinden bevor du sie umpartitionierst |
| 10:30 Barbarossa | hmm |
| 10:32 eartoast | stockholm: bei softupdates wird nicht partitioniert |
| 10:33 eartoast | deswegen ja 'soft' |
| 10:34 eartoast | nochmal im channel: grub hat die information, welche disk vom bios gebootet wird, in der device.map |
| 10:35 eartoast | irgendwie muss es also diese infos vom bios bekommen... |
| 10:36 stockholm | eartoast: jetzt wird mir gesagt dass die device.map meistens falsch ist und nicht benutzt wird |
| 10:36 h01ger | eartoast, ja, aber. reinstall ist auch ein sinnvolles features. softupdate erschlaegt viel, abner nicht alles |
| 10:36 h01ger | eartoast, den gedanken hatte ich gestern auch schon :) |
| 10:37 eartoast | stockholm: na toll ;( |
| 10:38 stockholm | ich habe scheinbar ein problem mich verständlich zu machen, kann jemand #grub joinen auf freenode? |
| 10:41 stockholm | eartoast: wenn du ein laufendes, funktionierendes system hast weisst du auch welche partition welche ist und kannst notfallst in ein maintainancesystem umsteigen und soche sachen machen |
| Action: Mrfai versteht zur zeit nur Bahnhof. ISt aber egal, ich geh jetzt erstmal essen |
| 10:43 eartoast | die boot-info ist da aber auch nicht mehr so richtig da. |
| 10:43 eartoast | du weisst halt nur, was dein momentanes root ist |
| 10:43 stockholm | eartoast: dann brauchst du aber die bootinfo auch nicht mehr |
| 10:44 eartoast | ob es andere roots auf anderen platten gibt, weisst du nicht |
| 10:45 stockholm | nee, aber die kümmern mich ja auch garnicht |
| 10:52 h01ger | aber welche ist die richtige? |
| 10:55 stockholm | h01ger: die die benutzt wird |
| 10:58 h01ger | bei reinstall bootest du via pxe |
| 10:58 h01ger | oder von usbstick aber nicht von platte |
| 11:17 stockholm | h01ger: ja, zur zeit. aber das sage ich ja gerade: an muss das vielleicht ändern? |
| 11:22 h01ger | ? |
| 11:23 h01ger | bei reinstall kannst/willst du nicht (unbedingt) von platte booten |
| 11:23 h01ger | sonst kannst du wirklich softupdate machen |
| kugg (~jerkeby@89.130.6.136) joined #fai. |
| allee (~ach@dialin-145-254-253-190.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| allee (~ach@dialin-212-144-128-092.pools.arcor-ip.net) joined #fai. |
| 12:11 oz_ | haha |
| Action: oz_ installiert nen ubuntu kernel in die nfsroot |
| 12:12 oz_ | *würg* |
| 12:14 allee | oz_: ? |
| meandtheshell (~markus@85.127.114.128) joined #fai. |
| meandtheshell (~markus@85.127.114.128) left irc: Quit: Leaving. |
| meandtheshell (~markus@85.127.114.128) joined #fai. |
| therziger (~CXO_Tobi@195.122.130.2) joined #fai. |
| therziger (~CXO_Tobi@195.122.130.2) left irc: Remote host closed the connection |
| woffs (~woffs@osiris.taz.de) joined #fai. |
| MT (~MT@dove.informatik.tu-muenchen.de) left irc: Remote host closed the connection |
| MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. |
| sobersabre (~bilbo@89-138-78-195.bb.netvision.net.il) joined #fai. |
| 13:16 sobersabre | hi all. |
| 13:17 sobersabre | can you please take a look at http://pastebin.ca/741435 and tell me ... why would a 2 disk (hda,hdb) client fail to install with errors about setup_harddisks |
| 13:18 sobersabre | the main disk is 80 GB, 1st partition on it is 52GB. |
| 13:18 sobersabre | the second disk is not to be touched at all. |
| 13:18 MT | preserve partition 1 is not valid syntax |
| 13:19 MT | you should say |
| 13:19 MT | primary - preserve |
| 13:20 sobersabre | oh... I think I copy-pasted it from somewhere. |
| 13:20 sobersabre | thanks! |
| 13:20 sobersabre | but ... there may be more than 1 primary partitions on a disk ( it's 4 ) |
| 13:21 Mrfai | primary - preserve1 |
| 13:21 Mrfai | this is th correct line |
| 13:21 sobersabre | OK. |
| 13:21 sobersabre | and: |
| 13:22 sobersabre | primary /media/C-ntfs preserve1 |
| 13:22 sobersabre | if I want it mounted ? |
| 13:23 sobersabre | probably /media/C would be shorter. |
| 13:25 Mrfai | yep |
| 13:25 sobersabre | I have a ... principle question: is it possible ( if yes, then how ) to remotely reconfigure BIOS/EFI setting to make a machine boot via PXE, IBA, or other means ? |
| 13:25 sobersabre | I understand that if a machine supports management network, like IBM server, it is not so complicated. |
| 13:26 Mrfai | or: primary /media/C preserve1 ; ntfs |
| 13:26 sobersabre | is there a way to interface the bios/efi on a currently running muchine ? |
| 13:26 sobersabre | Mrfai: thanks, fixing. |
| CXO_Tobi (~CXO_Tobi@195.122.130.2) joined #fai. |
| 13:27 Mrfai | sobersabre: no, you can't read or write the BIOS from linux in a way that will be portable. |
| 13:27 sobersabre | I haven't asked it to be portable. |
| 13:27 sobersabre | :) |
| 13:27 sobersabre | I mean is there a C/python whatever library or even application to do such things ? |
| 13:28 Mrfai | Therefore I suggest to set the BIOS to always do PXR boot. Then you can use fai-chboot to tell your machine if it should startinstalltion or boot from locl disk |
| 13:28 sobersabre | is its usage in man fai-chboot ? |
| Action: sobersabre is reading, thanks |
| 13:35 Mrfai | sobersabre: IIRC there's a nvram kernel module. |
| 13:36 Mrfai | sobersabre: http://www.cs.usfca.edu/~cruse/cs635/cmos.c, but IMO you will never write to those hex addresses. |
| 13:36 sobersabre | you mean in the ring the os is in ? |
| 13:36 sobersabre | you mean the cpu will sigill me ? |
| 13:36 sobersabre | or sigsegv ? |
| 13:37 sobersabre | I think fai-chboot is the way I need. |
| 13:37 Mrfai | IMO you are able to write to those locations, but if a minor different BIOS version will just change the location of some information, you are lost. |
| 13:37 Mrfai | :-) |
| 13:38 Mrfai | fai-chbbot is your friend, not writing to BIOS addresses. |
| 13:38 sobersabre | Mrfai: this means if I play with it, I better not update BIOS |
| 13:38 sobersabre | true :) |
| 13:38 sobersabre | All I need now is a way to remotely reboot windows machine. |
| 13:38 sobersabre | I think I'll do it via nagios. |
| 13:39 sobersabre | Mrfai: do you use windows too in your labs ? |
| 13:43 Mrfai | win.... what? Never heared of this software. |
| Action: Mrfai is really happy not to use such software. |
| Lin (~igor@200.179.57.57) joined #fai. |
| Action: Mrfai only have installed tow notbooks dualboot |
| 14:00 sobersabre | Mrfai: I wish I was as happy. |
| 14:00 sobersabre | Fenstern ... |
| alleeHol (~ach@dialin-145-254-255-197.pools.arcor-ip.net) joined #fai. |
| 14:01 sobersabre | Do you guys call it Fenstern in Germany ? |
| 14:01 sobersabre | :) |
| 14:02 sobersabre | italians may try: finestre di micromorbido |
| allee (~ach@dialin-212-144-128-092.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| 14:08 woffs | Hi! Is there already a patch for fai-chboot to correctly handle PXE Config Files with the MAC-DHCP-ID (instead of the IP) as filename? Like de-ad-be-ef-co-ff-ee |
| 14:11 eartoast | sobersabre: you can actually use 'ipmitool chassis bootdev $something' |
| 14:11 eartoast | if you have IMPI-hardware, of course |
| 14:11 Mrfai | woffs: yes, I have a patch for this, but it's not yet included. IIRC the patch is for an older version and changes more things in fai-chboot. So I need time to review this patch. |
| 14:16 woffs | Mrfai: can I have it, though, please? :) |
| kugg (~jerkeby@89.130.6.136) left irc: Read error: Connection reset by peer |
| 14:22 Mrfai | woffs: I've send you an email containing the diff. You may also need to patch FAI, because the client calls fai-chboot from remote using its hostname. You have to change that to use the MAC address of the client. |
| ErKa (keryell@m5.wifi.enstb.org) left irc: Ping timeout: 480 seconds |
| 14:51 woffs | Mrfai: thank you! in my environment, I do not call fai-chboot from the clients yet because (a) I dont really need it, and (b) security considerations. No clusters here. |
| baldy_ (baldy@brain.servercrew.de) left irc: Ping timeout: 480 seconds |
| 15:26 sobersabre | eartoast: how can I check if my hw is impi capable ? |
| ErKa (keryell@keryell.pck.nerim.net) joined #fai. |
| 15:27 sobersabre | I think the presence of "service processor" on a server is a good sign... but I remember something called "IMPI bus" |
| 15:27 sobersabre | so service processor only shouldn't be enough. |
| 15:27 sobersabre | lemme read. |
| baldy (baldy@brain.servercrew.de) joined #fai. |
| 15:30 oz_ | just a quick question... |
| 15:30 oz_ | is /srv/fai/config mounted into $target during install? |
| 15:31 sobersabre | yes |
| 15:31 sobersabre | /var/lib/fai/config |
| 15:31 sobersabre | you can see it if you issue "mount" |
| 15:32 sobersabre | hmm..it's not into $target, but on the client. |
| 15:32 sobersabre | isn't $target is /tmp/target ? |
| 15:32 oz_ | okay, thx |
| 15:32 sobersabre | oz_: do you simply want to see on the install client the config directory ? |
| 15:33 sobersabre | then I've answered your q. |
| 15:34 oz_ | i wanna install some custom kernel |
| 15:34 sobersabre | is it in the repository of your site ? |
| 15:34 oz_ | nope |
| 15:34 sobersabre | install or boot from it during the installation ? |
| 15:34 oz_ | just installing it |
| 15:35 sobersabre | and has it been prepared with make_kpkg ? |
| 15:36 oz_ | the pkg does exist, yes |
| 15:36 Mrfai | oz_: you should created a local repository of your own packages. kernel is also only a deb package |
| 15:37 oz_ | uhm, I always had problem with this... |
| 15:39 sobersabre | oz_: it is VERY easy to do, man. |
| 15:40 sobersabre | the apt howto is great. |
| 15:40 oz_ | sobersabre: I did not work when I tried it last time. |
| 15:40 oz_ | sI did/it did/ |
| 15:55 sobersabre | oz_: what exactly did not work ? ;) |
| meandtheshell (~markus@85.127.114.128) left irc: Quit: Leaving. |
| MT (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds |
| 15:59 sobersabre | eartoast: I think IPMI requires special BMC network ( usually comes as 10/100 Mb/s NIC ) |
| 16:00 sobersabre | so, the blades usually do have it, the IBM/HP rackmounted machines also do have it, but not boxes. |
| 16:00 sobersabre | I wonder about mac minis. |
| 16:01 sobersabre | I do read about guys trying to make a cluster of Mac Minis, but ... how well do they live on the long run ? |
| 16:01 sobersabre | has anybody here tried it ? |
| allee_ (~ach@dialin-212-144-131-135.pools.arcor-ip.net) joined #fai. |
| 16:02 sobersabre | hmm.. this is all offtopic to FAI of course. |
| alleeHol (~ach@dialin-145-254-255-197.pools.arcor-ip.net) left irc: Read error: Connection reset by peer |
| 16:08 oz_ | sobersabre: dunno anymore what exactly happened |
| 16:08 oz_ | I have to do it again, when there is more time |
| 16:10 oz_ | sobersabre: IPMI is normally used with a special service network |
| 16:15 sobersabre | yep, I have it on blade center, but the workstations... it's not so practical to have 2 cables. |
| 16:16 sobersabre | and the workstations are partially contributed sh1te. |
| 16:16 sobersabre | outdated, and not highend. |
| 16:19 allee_ | sobersabre: AFAIR there a IPMI daemon for linux. Of course you can do anything with the daemon before the Workstations are running ;) |
| 16:20 allee_ | s/can/can not/ |
| 16:23 sobersabre | allee_: you mean the daemon runs on the same cpu with the main machine, so I could add them to misc. management tools.. nice. |
| 16:24 sobersabre | how does the wake on lan work ? |
| 16:34 Mrfai | http://en.wikipedia.org/wiki/Wake_on_lan |
| Mrfai (~lange@kueppers.informatik.uni-koeln.de) left irc: Quit: leaving |
| woffs (~woffs@osiris.taz.de) left irc: Quit: Leaving. |
| MT (~MT@dove.informatik.tu-muenchen.de) joined #fai. |
| alleeHol (~ach@dialin-212-144-130-119.pools.arcor-ip.net) joined #fai. |
| allee_ (~ach@dialin-212-144-131-135.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| 18:28 sobersabre | byebye |
| sobersabre (~bilbo@89-138-78-195.bb.netvision.net.il) left irc: Quit: leaving |
| alleeHol (~ach@dialin-212-144-130-119.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| 18:42 stockholm | MT: i think you need to restructure your controlflow |
| 18:42 MT | hmm, why? |
| 18:43 stockholm | right now you pile all lvm and all raid and all partition commonads on top of each other |
| 18:43 stockholm | but you sould really pile not by lvm, raid and partioins |
| 18:44 MT | what do you suggest instead? |
| 18:44 stockholm | but first after remove (and within remove after lvm, raid, and partitions) |
| 18:44 stockholm | and then after create (and within create after partitions, raid and lvm) |
| 18:44 stockholm | really bad net here |
| 18:45 MT | in fact, when I did that part of the code I was pretty concerned about the flow and not really happy with the way I did it |
| 18:45 MT | but I was then running out of time, so I kept it the way it was and now is |
| alleeHol (~ach@dialin-212-144-130-119.pools.arcor-ip.net) joined #fai. |
| 18:45 MT | you're definitely right that is not really cool |
| 18:46 MT | in fact you can't do a RAID over LVM with the current code |
| 18:46 stockholm | yeah |
| 18:46 stockholm | and lvm deletion and creation fails in certain cases |
| 18:46 stockholm | because you first remove the partitions and THEN the lvm |
| 18:47 stockholm | instead of the other way around |
| 18:47 MT | hmm, yes, right |
| Lin (~igor@200.179.57.57) left irc: Remote host closed the connection |
| 18:47 MT | do you have any time to fix this? |
| 18:47 stockholm | my suggestion would be to break all those big funktions down into many smaller ones. |
| 18:47 MT | I won't find the time before the dev weekend |
| 18:47 MT | oh, that would be pretty cool |
| 18:48 stockholm | MT: i am unsure i want to overthrow your code |
| 18:48 stockholm | this would be a major thing to do |
| 18:48 MT | oh, feel free to do anything to it |
| 18:48 stockholm | so far i have done small fixes |
| 18:48 MT | I'm well aware that it's far from optimal |
| sep (~sep@89.130.6.136) left irc: Ping timeout: 480 seconds |
| 18:49 stockholm | MT: i will see what i can do |
| 18:49 stockholm | but i dont promise anything :-) |
| 18:49 MT | If you happen to throw away everything that's ok - as long as we end up with a better code |
| 18:49 MT | no need for promises, do whatever you can |
| 18:50 stockholm | we can use all the code, just the flow needs to be redirected... |
| 18:50 stockholm | yes |
| 18:50 MT | and what you can't do (or even those things you plan to do and probably will finish): please document it on the wiki page |
| 18:51 stockholm | i wanted to write a lengthly mail about the status |
| 18:51 MT | so we keep a TODO list and whoever has time may then feel free to hack on it following the TODO list |
| 18:51 stockholm | right |
| sep (~sep@89.130.6.136) joined #fai. |
| 18:52 MT | well, if you prefer mail that's fine as well, but the wiki page is probably easier to update in a shared manner |
| 18:52 stockholm | right |
| 18:52 stockholm | but wiki requires network |
| 18:52 stockholm | and networks sucks dead goats here |
| 18:52 MT | :-) |
| 18:53 MT | well, that's just the situation for the next few days |
| 18:53 stockholm | right |
| 18:53 MT | in fact, I won't have too much network access until tuesday anyway |
| 18:53 MT | as I'm travelling as well |
| 18:53 stockholm | i wanted to have a working fai setup after extremadura |
| 18:53 stockholm | and instead i got tricked into development |
| 18:53 sep | hihihih :) |
| 18:54 stockholm | lol |
| 18:54 MT | but community is prefers your hacking anyway :-) |
| 18:55 stockholm | right |
| Nick change: MT -> Guest2189 |
| MT (~MT@80.109.167.79) joined #fai. |
| 19:09 stockholm | MT: why dont we use this? http://search.cpan.org/~ckerner/Linux-LVM-0.13/LVM.pm |
| 19:11 MT | hmm, probably because I didn't know about it :-) |
| 19:13 MT | stockholm, I gotta leave now |
| 19:13 stockholm | it is not packaged yet |
| 19:13 MT | feel free to change anything ... |
| 19:13 stockholm | i just try to rope the perl group into packaging it |
| 19:14 MT | ok, that's the best way to do it |
| 19:14 MT | otherwise just add it to the SVN ... |
| MT (~MT@80.109.167.79) left irc: Quit: MT |
| Guest2189 (~MT@dove.informatik.tu-muenchen.de) left irc: Ping timeout: 480 seconds |
| sobersabre (~bilbo@89-138-78-195.bb.netvision.net.il) joined #fai. |
| sobersabre (bilbo@89-138-78-195.bb.netvision.net.il) left #fai. |
| alleeHol (~ach@dialin-212-144-130-119.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| alleeHol (~ach@dialin-212-144-130-090.pools.arcor-ip.net) joined #fai. |
| MT (~MT@M328P017.adsl.highway.telekom.at) joined #fai. |
| ErKa (keryell@keryell.pck.nerim.net) left irc: Ping timeout: 480 seconds |
| MT (~MT@M328P017.adsl.highway.telekom.at) left irc: Quit: MT |
| alleeHol (~ach@dialin-212-144-130-090.pools.arcor-ip.net) left irc: Ping timeout: 480 seconds |
| sep (~sep@89.130.6.136) left irc: Quit: Leaving |
| --- Sat Oct 20 2007 |