Re: ypbind not enabled after installation
On the ypclient: In /etc/default/nis set NISCLIENT=true In /etc/yp.conf set the yp (master and slave) server IP addresses On the yp master+slave servers: In /etc/yp.conf set the yp (master and slave) server IP addresses In /etc/ypserv.securenets list all IP addresses of all yp clients so the ypserver accepts yp requests Andreas
Re: Ubuntu 22.04: problem installing nvidia-driver-525
On 4/20/23 13:07, Thomas Lange wrote: On Thu, 20 Apr 2023 10:45:52 +0200, Andreas Sindermann said: > Hi all, > there seems to be a problem installing a Ubuntu 22.04 package that is > trying to call update-initramfs: > update-initramfs: Generating /boot/initrd.img-5.10.0-18-amd64 > W: missing /lib/modules/5.10.0-18-amd64 > W: Ensure all necessary drivers are built into the linux image! > depmod: ERROR: could not open directory /lib/modules/5.10.0-18-amd64: No > such file or directory Is 5.10.0-18-amd64 the kernel version, that is running during the installation? Or is it the ubuntu kernel version inside /target? Yes, it seems that 5.10.0-18-amd64 is indeed the Debian kernel running during the installation: root@jammysrv:/srv/fai/nfsroot/lib/modules# ls 5.10.0-18-amd64 My guess is that update-initramfs tries to build an initrd for the wrong kernel version. Can you just ignore this error, or does the installation aborts? The installation itself continues but the gpu is not identified when booting the installed Ubuntu 22.04: root@l96:~# lspci |grep -i nvi 01:00.0 VGA compatible controller: NVIDIA Corporation Device 2507 (rev a1) 01:00.1 Audio device: NVIDIA Corporation Device 228e (rev a1) It should create the initramfs in the Ubuntu environment, not in the Debian environment... Thanks! Andreas regards Thomas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
Ubuntu 22.04: problem installing nvidia-driver-525
Hi all, there seems to be a problem installing a Ubuntu 22.04 package that is trying to call update-initramfs: root@jammysrv:~# cat /srv/fai/config/package_config/NVIDIACURRENT PACKAGES install linux-headers-generic nvidia-driver-525 #nvidia-cuda-toolkit Error messages in fai.log: Setting up nvidia-kernel-common-525 (525.105.17-0ubuntu0.22.04.1) ... update-initramfs: deferring update (trigger activated) update-initramfs: Generating /boot/initrd.img-5.10.0-18-amd64 W: missing /lib/modules/5.10.0-18-amd64 W: Ensure all necessary drivers are built into the linux image! depmod: ERROR: could not open directory /lib/modules/5.10.0-18-amd64: No such file or directory depmod: FATAL: could not search modules: No such file or directory cat: /var/tmp/mkinitramfs_Oa84zB/lib/modules/5.10.0-18-amd64/modules.builtin: No such file or directory Kind regards, Andreas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
Re: UEFI boot order, Re: Tip: Remote FAI install
Hi, this doesn't seem to be a trivial task as the state of the boot medium prior to the fai installation as well as the UEFI settings for the single network interface both can have all kinds of states, e.g.: (of course for production I'd disable unneeded UEFI setting, in my case all IPv6 and HTTP boot options, so just the IPv4 PXE boot option usually would be active leading to the IPv4 PXE having highest boot priority and then just installed HD/NVMe medium having the second boot priority) root@jammysrv:/srv/fai/config# cat hooks/partition.GRUB_EFI #! /bin/bash # prior to disk partioning collect the UEFI boot order efibootmgr --- root@jammysrv:/srv/fai/config# less /var/log/fai/remote-logs/l65/last/fai.log ---Output [...] Calling hook: partition.GRUB_EFI BootCurrent: 0001 Timeout: 1 seconds BootOrder: 0001,0002,0003,0004,,0005 Boot* ubuntu Boot0001* UEFI: PXE IP4 Intel(R) Ethernet Connection (2) I219-V Boot0002* UEFI: HTTP IP4 Intel(R) Ethernet Connection (2) I219-V Boot0003* UEFI: HTTP IP6 Intel(R) Ethernet Connection (2) I219-V Boot0004* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V Boot0005* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V partition.GRUB_EFI OK. [...] root@jammysrv:/srv/fai/config# cat scripts/GRUB_EFI/01-UEFI-Collect_BootOrder #! /bin/bash error=0; trap 'error=$(($?>$error?$?:$error))' ERR # save maximum error code efibootmgr -v exit $error ---Output = shell: GRUB_EFI/01-UEFI-Collect_BootOrder = BootCurrent: 0001 Timeout: 1 seconds BootOrder: 0001,0002,0003,0004,,0005 Boot* ubuntu HD(1,GPT,6205f216-7aa2-4531-8319-d7b77a00514d,0x800,0x10 )/File(\EFI\UBUNTU\GRUBX64.EFI) Boot0001* UEFI: PXE IP4 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci (0x1f,0x6)/MAC(b42e9987238b,0)/IPv4(0.0.0.00.0.0.0,0,0)..BO Boot0002* UEFI: HTTP IP4 Intel(R) Ethernet Connection (2) I219-V PciRoot( 0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv4(0.0.0.00.0.0.0,0,0)/Uri()..BO Boot0003* UEFI: HTTP IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot( 0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)/Uri()..BO Boot0004* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)..BO Boot0005* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)..BO GRUB_EFI/01-UEFI-Collect_BootOrder OK. = shell: GRUB_EFI/10-setup = ainsl: appending to /target/etc/default/grub: GRUB_DISABLE_OS_PROBER=true Installing for x86_64-efi platform. Installation finished. No error reported. Grub installed on /dev/nvme0n1 = (hostdisk//dev/nvme0n1) Sourcing file `/etc/default/grub' Sourcing file `/etc/default/grub.d/init-select.cfg' Generating grub configuration file ... Found linux image: /boot/vmlinuz-5.15.0-58-generic Found initrd image: /boot/initrd.img-5.15.0-58-generic Memtest86+ needs a 16-bit boot, that is not available on EFI, exiting Warning: os-prober will not be executed to detect other bootable partitions. Systems on them will not be added to the GRUB boot configuration. Check GRUB_DISABLE_OS_PROBER documentation entry. Adding boot menu entry for UEFI Firmware Settings ... done GRUB_EFI/10-setupOK. root@jammysrv:/srv/fai/config# cat scripts/GRUB_EFI/90_UEFI_Adjust_BootOrder_pre #! /bin/bash error=0; trap 'error=$(($?>$error?$?:$error))' ERR # save maximum error code efibootmgr -v exit $error ---Output = shell: GRUB_EFI/90_UEFI_Adjust_BootOrder_pre = BootCurrent: 0001 Timeout: 1 seconds BootOrder: ,0001,0002,0003,0004,0005 Boot* ubuntu HD(1,GPT,ac900736-afc3-4b46-8bb5-ce1c1b71c9a8,0x800,0x10)/File(\EFI\ubuntu\grubx64.efi) Boot0001* UEFI: PXE IP4 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv4(0.0.0.00.0.0.0,0,0)..BO Boot0002* UEFI: HTTP IP4 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv4(0.0.0.00.0.0.0,0,0)/Uri()..BO Boot0003* UEFI: HTTP IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)/Uri()..BO Boot0004* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)..BO Boot0005* UEFI: PXE IP6 Intel(R) Ethernet Connection (2) I219-V PciRoot(0x0)/Pci(0x1f,0x6)/MAC(b42e9987238b,0)/IPv6([::]:<->[::]:,0,0)..BO GRUB_EFI/90_UEFI_Adjust_BootOrder_pre OK. root@jammysrv:/srv/fai/config# cat scripts/GRUB_EFI/91_UEFI_Adjust_BootOrder #! /bin/bash error=0; trap 'error=$(($?>$error?$?:$error))' ERR # save maximum error code # FAI List 19.01.2023, 21:10h (Thomas Lange) #Hi, # #I found this code that move the first boot entry (which is expected to #be the new entry after an installation) to the end of the boot list. #
Re: UEFI boot order, Re: Tip: Remote FAI install
Hi all, I'm wondering whether this issue concerning bad/wrong UEFI boot order after a fai installation already was resolved in the meantime (since September, 2021)? https://lists.uni-koeln.de/pipermail/linux-fai/2021-September/012770.html The idea was to e.g. add a few lines in config/scripts/GRUB_EFI/10-setup with efibootmgr commands (possibly before and) after the grub-install commands to correct the changed UEFI boot order. Currently I'm installing new desktops with uninitialized ssds and after the reboot I find the PXE boot option with a lower priority than the installed OS which for my environment is quite inconvenient... Best regards, Andreas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
Re: FAI server on Ubuntu
On 11/08/2017 09:27 AM, Thomas Lange wrote: > On IRC we had the question, if FAI can be installed on Ubuntu. Does > anyone has a FAI server running on Ubuntu Xenial? I remember that > there were major problems creating the nfsroot on a Ubuntu system, > because of dracut and upstart in the past. But I'm not sure if these > problems still apply or not. Any help appreciated. > The fai-server package available with Ubuntu 16.04 LTS can be used. Nevertheless it will create a Debian-nfsroot environment which then can install an Ubuntu system on the individual clients. For this last step you have to create a Ubuntu base file system config/basefiles/CLASSNAME.tar.gz with debootstrap. The details for this debootstrap command can be found somewhere in the fai-setup script. Best regards Andreas
multiarch issue with FAI
Dear all, I'm trying to install a 64-bit Ubuntu Linux. To be able to use acroread and other 32-bit software I'm trying to setup a multi arch environment using a fai hook to install the i386 specific libraries: root@faiserver:/srv/fai/config/hooks# cat updatebase.AMD64 #!/bin/bash echo Adding i386 multi-arch $ROOTCMD dpkg --add-architecture i386 root@faiserver:/srv/fai/config/package_config# cat AMD64 PACKAGES aptitude libgtk2.0-0:i386 libnss3-1d:i386 libnspr4-0d:i386 lib32nss-mdns* libxml2:i386 libxslt1.1:i386 libstdc++6:i386 Unfortunately during the fai installation process the hook is executed at a very early stage so that the dpkg package is not yet installed on the client. In the end the i386 packages mentioned above are not installed: [... fai.log ...] Calling task_prepareapt Calling hook: updatebase.AMD64 Adding i386 multi-arch updatebase.AMD64 OK. Calling task_updatebase Updating base /target/usr/sbin/dpkg-divert not available. Skipping. /target/usr/sbin/dpkg-divert not available. Skipping. [...] Of course I simply could introduce a script in the fai scripts-directory, which would be executed at the end of the installation but on the tips'n tricks web page the solution with the hook is presented: http://wiki.fai-project.org/wiki/Tips_and_tricks So I'm wondering whether there would be a more convenient solution... Thanks and best wishes Andreas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
[SOLVED] Re: FAI scp problem writing logfiles
Tobias Unsleber writes: Hi Andreas, after checking Holger's advice you may test this too: a) Check the system/auth log files of the fai-server regarding any ssh-authentication issues(e. g. authorizedkeys file world writable) b) Check to ssh from the installed client (When the client machine waiting is waiting for Return to reboot) to the faiserver using the fai LOGUSER c) check the local fai-logs at the waiting client (/var/log/fai I think) Simple solution: For whatever reasons the fai account on the faiserver was not owned by user LOGUSER but instead by root (althought the ~LOGUSER/.ssh directory was correctly owned by LOGUSER). Of course this leads to permission issues when trying to remotely fai-chboot on or to scp files to the faiserver. By chown'ing the home directory to LOGUSER everything now works fine: root@faiserver:~# ls -ld ~fai drwxr-xr-x 4 fai root 4096 Aug 24 14:35 /var/log/fai This happened with version: root@faiserver:~# apt-cache showpkg fai-server Package: fai-server Versions: 3.4.8ubuntu5 (/var/lib/apt/lists/de.archive.ubuntu.com_ubuntu_dists_trusty_universe_binary-amd64_Packages) (/var/lib/dpkg/status) Thanks for your hints and best wishes Andreas
FAI scp problem writing logfiles
Dear all, probably it's my stupidity but with a new FAI installation server I have some scp problem of this type at the very end of the installation process: [...] ERRORS found in log files. See /tmp/fai/error.log savelog.LAST.source OK Calling task_savelog Save log file via ssh to x...@xxx.xxx.xxx.xxx:yyy//install-20150820_14 lost connection Calling task_faiend Press RETURN to reboot. [...] So copying of the logfiles to the FAI server fails. Any ideas? $LOGUSER/.ssh/authorized_keys does exist. An interactive ssh session to my personal account on the FAI server does work properly. But copying via scp to the $LOGUSER account fails: root@faiclient:/# ll abc -rw-r--r-- 1 root root 5 Aug 20 15:58 abc root@faiclient:/# scp abc fai@trusty64srv: lost connection root@faiclient:/# Thanks for any feedback. Best Andreas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
Ubuntu natty setup-storage questions
Again me with three issues I asked some time ago already... a) setup-storage leads to a different partition table than setup_harddisks, although the disk_config file only changed in syntax. b) I'm currently using two classes NO_SCRATCH (for a standard partition layout that all of the FAI clients are using) and SCRATCH (to preserve a possibly existing /scratch partition that I created by hand earlier on some of the FAI clients, but not on all of them). b 1) How could theses classes NO_SCRATCH and SCRATCH be merged in one single class? Is it possible at all? b 2) There seems to be a problem with setup-storage not being able to preserve the /scratch partition in class SCRATCH, see below. Here the long story: Some of my FAI clients have a local /scratch partition on /dev/sda3, some other FAI clients don't have such a partition. Up to now I created these partitions by hand. After a reinstallation via FAI (setup_harddisks) with Ubuntu 8.04 I had to restore the partition table by hand. Ubuntu 8.04 setup_harddisk disk_config file: # type mountpoint size in mb [mount options] [;extra options] disk_config disk1 primary / 1 rw,errors=remount-ro ; -j ext3 logical swap 2000 rw The disk_config setup results in this partition table: root@l28:~# fdisk /dev/sda Disk /dev/sda: 250.0 GB, 250059350016 bytes 255 heads, 63 sectors/track, 30401 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0x0001e7ff Device Boot Start End Blocks Id System /dev/sda1 * 1127510241406 83 Linux /dev/sda212761530 2048287+ 5 Extended /dev/sda512761530 2048256 82 Linux swap / Solaris After creating/restoring /scratch by hand: (depending on the capacity of the harddisk its size varies, but it always starts at cylinder 1531!) root@l28:~# fdisk /dev/sda Disk /dev/sda: 250.0 GB, 250059350016 bytes 255 heads, 63 sectors/track, 30401 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0x0001e7ff Device Boot Start End Blocks Id System /dev/sda1 * 1127510241406 83 Linux /dev/sda212761530 2048287+ 5 Extended /dev/sda31531 30401 231906307+ 83 Linux /dev/sda512761530 2048256 82 Linux swap / Solaris Currently, we are migrating to Ubuntu natty 11.04. The new setup-storage script leads to a slightly different layout, although the new disk_config file looks pretty much the same as before: root@faisrv:/srv/fai/config/disk_config# cat NO_SCRATCH # type mountpoint size fs type mount options misc options disk_config disk1 disklabel:msdos bootable:1 fstabkey:uuid primary /1ext3 rw,noatime,errors=remount-ro logical swap 2000swap rw root@l41:~# fdisk /dev/sda Disk /dev/sda: 250.1 GB, 250058268160 bytes 255 heads, 63 sectors/track, 30401 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x00050434 Device Boot Start End Blocks Id System /dev/sda1 * 1127410233373+ 83 Linux /dev/sda212751528 2040255f W95 Ext'd (LBA) /dev/sda512751528 2040223+ 82 Linux swap / Solaris For whatever reasons with Ubuntu 11.04 (setup-storage) /dev/sda1 ends at cylinder 1274 instead of 1275, additionally /dev/sda2 ends at cylinder 1528 instead of 1530 as with Ubuntu 8.04 (setup_harddisks). Is there a way that setup-storage would lead to the same layout like before with setup_harddisks? For the fai clients with a /scratch partition I created a similar disk_config file: # type mountpoint size fs type mount options misc options disk_config disk1 disklabel:msdos bootable:1 fstabkey:uuid preserve_always:3 primary /1ext3 rw,noatime,errors=remount-ro logical swap 2000swap rw primary /scratch 0- ext3 rw I'd like to merge these two classes if possible. Is there a way to preserve a partition of unknown size and unknown starting point, but only if it does exist already? I'm doing initial installs (fai-chboot -IvBk nouveau.modeset=0 fai_client), but still would like to preserve /scratch. I think 'preserve_lazy' is doing more than just preserving the partition. And, secondly, the disk_config file leads to an error: root@faisrv:/home/fai/l28/last# more format.log Starting setup-storage 1.3 Using config file: /var/lib/fai/config/disk_config/SCRATCH Executing: parted -s /dev/sda unit TiB print Executing: parted -s /dev/sda unit B print free Executing: parted -s /dev/sda unit chs print free Creating directory /var/lock/lvm Finding all volume groups No volume groups found Executing: mdadm --examine --scan
Re: Ubuntu natty setup-storage questions
Thomas Neumann writes: b) I'm currently using two classes NO_SCRATCH (for a standard partition layout that all of the FAI clients are using) and SCRATCH (to preserve a possibly existing /scratch partition that I created by hand earlier on some of the FAI clients, but not on all of them). b 1) How could theses classes NO_SCRATCH and SCRATCH be merged in one single class? Is it possible at all? What is your goal? a) Do you manually decide which class applies for each host and want to automate/simplify the decision process? b) Do you want to clean up/consolidate your config space and want to handle /scratch as a special case of NO_SCRATCH? The answers would be different. a) would hint at some decision logic in class/ which automatically announces the correct class. b) would require either modifying your layout-files on the fly or change the way disk_config works. Everything's possible. We just need to find the right hammer for your nail. ;) Currently I'm using the following script to decide whether a client does already have a scratch-partition or not: root$ cat FAI/class/30-detect-scratch-partitions #! /bin/bash # detect whether a scratch partition /dev/sda3 already exists size=`fdisk -s /dev/sda3 2/dev/null` if [ ! -z $size ]; then echo SCRATCH ; else echo NO_SCRATCH; fi exit 0 I'd prefer a single setup-storage layout file that automagically detects and preserves an already existing partition. It seems that the current setup-storage a) changed behaviour concerning cylinder boundaries compared to setup_harddisks b) offers different methods how to preserve partitions, but none of them seems to fit my case (or I misunderstood the manual, that might be the case, of course...). c) gives a simple error, which should not happen. Or my layout file has an error that I don't understand... Andreas
Ubuntu natty setup-storage question
An issue with setup-storage in Ubuntu 11.04 (natty) occurs: root@faisrv:/srv/fai/config/disk_config# cat ~fai/l00/last/format.log Starting setup-storage 1.3 Using config file: /var/lib/fai/config/disk_config/SCRATCH Executing: parted -s /dev/sda unit TiB print Executing: parted -s /dev/sda unit B print free Executing: parted -s /dev/sda unit chs print free Creating directory /var/lock/lvm Finding all volume groups No volume groups found Executing: mdadm --examine --scan --verbose -c partitions /dev/sda3 will be preserved Executing: vgchange -a n Executing: parted -s /dev/sda mklabel msdos Executing: parted -s /dev/sda mkpart primary ext3 32256B 10479006719B Executing: parted -s /dev/sda mkpart extended 500105249280B 502194470399B Command had non-zero exit code Manually on the fai client: root@l00:~# parted -s /dev/sda mklabel msdos root@l00:~# parted -s /dev/sda mkpart primary ext3 32256B 10479006719B Warning: The resulting partition is not properly aligned for best performance. root@l00:~# parted -s /dev/sda mkpart extended 500105249280B 502194470399B Error: The location 502194470399B is outside of the device /dev/sda. root@l00:~# parted -s /dev/sda unit B print free Model: ATA ST3500514NS (scsi) Disk /dev/sda: 500107862016B Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start EndSize Type File system Flags 1 32256B10479006719B 10478974464B primary ext3 10479006720B 500107862015B 489628855296B Free Space So 502194470399B seems to be really outside of the device (swap-partition). Here the disk_config/SCRATCH file: root@faisrv:/srv/fai/config/disk_config# cat SCRATCH # example of new config file for setup-storage # # type mountpoint size fs type mount options misc options disk_config disk1 disklabel:msdos bootable:1 fstabkey:uuid preserve_always:3 primary /1ext3 rw,noatime,errors=remount-ro logical swap 2000swap rw primary /scratch 0- ext3 rw I.e. if already existing, the /scratch partition is supposed to be preserved, otherwise it should be created. Any idea? And why does parted want to start the first partition at 32256B? (bad performance?) Best wishes Andreas
Question about setup-storage
Hello, I have a question about the config file for setup-storage. Using the table disk_config disk1 primary / 1 rw,errors=remount-ro ; -j ext3 logical swap 2000 rw the old *setup_harddisk* created the following partition table: Device Boot Start End Blocks Id System /dev/sda1 * 1127510241406 83 Linux /dev/sda212761530 2048287+ 5 Extended /dev/sda512761530 2048256 82 Linux swap / Solaris Additionally on some of our machines a /scratch partition starting at cylinder 1531 was established: /dev/sda31531 30401 231906307+ 83 Linux In contrast to that the new setup-storage command creates a somewhat different partition table. disk_config disk1 disklabel:msdos bootable:1 fstabkey:uuid primary /1ext3 rw,noatime,errors=remount-ro logical swap 2000swap rw leads to partitions which are 1 cylinder shorter than before: Device Boot Start End Blocks Id System /dev/sda1 * 1127410233373+ 83 Linux /dev/sda212751528 2040255f W95 Ext'd (LBA) /dev/sda512751528 2040223+ 82 Linux swap / Solaris How should the configuration file for setup-storage look like to reproduce the original partition table? The man page doesn't mention 'c' for cylinders as size specification. Thanks in advance Andreas
Re: Question about setup-storage
Michael Tautschnig writes: Hi Andreas, [...] the old *setup_harddisk* created the following partition table: =20 Device Boot Start End Blocks Id System /dev/sda1 * 1127510241406 83 Linux /dev/sda212761530 2048287+ 5 Extended /dev/sda512761530 2048256 82 Linux swap / Sola= ris =20 Additionally on some of our machines a /scratch partition starting at cylinder 1531 was established: =20 /dev/sda31531 30401 231906307+ 83 Linux =20 =20 =20 In contrast to that the new setup-storage command creates a somewhat different partition table. =20 disk_config disk1 disklabel:msdos bootable:1 fstabkey:uuid primary /1ext3 rw,noatime,errors=3Dremount-ro logical swap 2000swap rw =20 leads to partitions which are 1 cylinder shorter than before: =20 Device Boot Start End Blocks Id System /dev/sda1 * 1127410233373+ 83 Linux /dev/sda212751528 2040255f W95 Ext'd (LBA) /dev/sda512751528 2040223+ 82 Linux swap / Sola= ris =20 =20 How should the configuration file for setup-storage look like to reproduce the original partition table? =20 The man page doesn't mention 'c' for cylinders as size specification. =20 Cylinders aren't really useful these days as the numbers you get don't rela= te to anything really physically existing. Therefore I don't see much use in implementing this unit of measurement. If you need a really precise specification you may go to the level of bytes (use parted -s /dev/bla -s u= nit B print to obtain that information from one of the existing systems). As setup-storage has been implemented from scratch, the details of computin= g the size, which involve rounding to boundaries, may show subtle differences with respect to the results of setup_harddisks. But, just curious, why does that small difference matter for you? These scratch partitions should be preserved if possible... I'll use parted -s instead of setup-storage... Thanks for your answer. Andreas Best regards, Michael --82I3+IH0IqGh5yIs Content-Type: application/pgp-signature -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.11 (Darwin) iQIcBAEBCAAGBQJNgO+qAAoJEKjvEonjOGdBNLoP/ifFOaxZfRDOQV0JXst5DyxJ l8r3F1NfKXlhmbNNql5Ti43x3GeCmxsooPBmk5xQ3uH9vMTqqYHXOyc/Is1kLrc2 74vHw/iaSb1lBFNDFrKBlLIvJjAIp9QODTCqfx+KReOtBDmbxtCc9lZ7adLtg0M+ Wyvb/OBDitp+oSqBdXHbfEkbL4IExEsgEr7YDMvi+rHNL50Rcu7i89nRTZk6O13V ZW/2G+zhHGuDy6jSznE4KM6t4BMz/4RdMW+0B8SvqPcHjRvMWLcwpyIsocufMfHf pSehZdb/CZAzj7koExg3gJXBijLSOr4B3cQkjVBnMxfy3EN78PmvFhqJ5EEhrJCa jFGVAlH7lsZRcj26Hu0SWXQcF30GEb/HU4CYu+29+h80ox+0kpuX3L8ROGQcydBg M767Cn0CyNDbs3ManAfABybsVmIyQ+8p7FECvlj+tT/wSYsor1eiwUuhSk5i1nOw q3Oqb+0WxE2DYbJ7PHtO87dfuHKgLX7Znxg7OI4fnSt4j7rmZ9chBGMagFqJ/mLv 1XSbi5P1+3gPI6UQl6hgh1BEYGyf1EJsR+oWY7CvBnsv98D8y9pLarFaFS7GcaCs HoM4XcuBk9ublFqSDzqzULuvPsrxDpGLk8608dW7Szh9+fT3eYWOlJxZKOMNb1a7 xnGE0J+byy220aZaThfX =wWR0 -END PGP SIGNATURE- --82I3+IH0IqGh5yIs--
Ubuntu natty: debootstrap doesn't install aptitude
root@faisrv:~# lsb_release -rd Description:Ubuntu natty (development branch) Release:11.04 The FAI /usr/sbin/make-fai-nfsroot script creates a chroot environment ('nfsroot') using the debootstrap command. At least with Ubuntu 11.04 Alpha 2 the debootstrap command no longer installs the 'aptitude' package by default in this nfsroot environment. As 'aptitude' is missing, the nfsroot environment stays incomplete and cannot be used for the further installation process. To resolve this issue, the natty fai-server package needs to be modified this way: In /etc/apt/make-fai-nfsroot.conf the last line should be changed to: FAI_DEBOOTSTRAP_OPTS=--exclude=info --include=aptitude Andreas
Re: Ubuntu natty: debootstrap doesn't install aptitude
Stephan Adig writes: Hi Andreas, On Mon, 2011-02-07 at 14:32 +0100, Andreas Sindermann wrote: root@faisrv:~# lsb_release -rd Description:Ubuntu natty (development branch) Release:11.04 The FAI /usr/sbin/make-fai-nfsroot script creates a chroot environment ('nfsroot') using the debootstrap command. At least with Ubuntu 11.04 Alpha 2 the debootstrap command no longer installs the 'aptitude' package by default in this nfsroot environment. As 'aptitude' is missing, the nfsroot environment stays incomplete and cannot be used for the further installation process. To resolve this issue, the natty fai-server package needs to be modified this way: Well, I would like to replace aptitude with apt-get in /etc/fai/NFSROOT can you test this? Replace in /etc/fai/NFSROOT the name aptitude with install Rational: As Ubuntu still stays with apt-get, we should at least try to use the installed by default utilities. I don't know the status of aptitude inside Ubuntu (I don't use aptitude myself), but having Michael Vogt on board of the Canonical ship, I think it would better to stay with apt-get and not aptitude. This is my personal opinion, and this is Ubuntu only. This is someting more 'political' I would say. In terms of keeping the original package as 'original' as possible one better should resolve the aptitude dependency as described above, as Thomas seems to like aptitude... :-) On the other hand it might be a good idea to generally switch to apt-get. But my suggestion would be to do this in the Debian packages. Some time later the Ubuntu packages also would not rely on aptitude any longer as an automatic outcome. Regards Andreas
fai server on Ubuntu lucid?
Hi, did anyone already try to install a fai server on a Ubuntu Lucid platform? As far as I understand Lucid clients have been successfully installed from a server on Ubuntu Karmic. Unfortunately for some reason that I don't understand at the moment a Ubuntu Karmic installation fails on my hardware platform... Andreas -- Dr. Andreas Sindermann fon: +49 (221) 470-4201 Institut fuer Theoretische Physikfax: +49 (221) 470-5159 Universitaet zu Koeln Zuelpicher Str. 77 mailto:sin...@thp.uni-koeln.de D-50937 Koeln, Germany http://www.thp.uni-koeln.de/~sinder
Re: NFS/dhcp/tftp on eth1, kernel search NFS it on eth0
Jean Spirat writes: Jean Spirat a écrit : Hello, All my tests works well but i have a failure when i try to install a host with two ethernet cards. The dhcp and tftp runs on the eth1 network and it works fine. But the issue is that the fai kernel when loaded bring not eth1 but eth0. So it can never connect to the nfs server.i tried to add ether=0,0,eth1 to the pxe config but this does not work as it seems the tg3 modules is not in kernel but comes from the initramfs. I do not see how i can change anything in the fai kernel or initd to make this works. Anyone know how to deal with this problem or know if i do something wrong ? regards, Jean. to add a little more explanation: - host boot ok - pxe boot get the ip from the dhcp - tftp load the kernel ok - kernel fail to get the NFS server as it brings only eth0 up Try disabling one of the ethernet interfaces in the BIOS. Hope this helps Andreas
Re: Bonus packages in fai-mirror
Russel Hill writes: We are building using fai-cd from a private package pool (a local subset of debian packages). Ordinarily, we build our private package pool by first building an ISO from more complete debian mirror. Then we post-process the access logs and built our private package pool from that list of packages. FWIW: This process has been working for us for a few years now. We appear to be getting many spurious packages on our fai-cd. For example, aspell-en gets pulled in by fai-mirror and installed on the target system. Absolutely nothing in the package poolsdepends on aspell-en. The only thing I can see is that aspell-en is Recommended by some packages. However, there are other recommended packages that don't get installed. aspell-en is only one example of a 'bonus' package. Our ISO has grown from 410MB with fai v3.1 to 560MB with fai v3.2.4. We can manually pull some of these spurious packages out but it's very tedious and time consuming. Does anybody have a clue what might be causing this? I'd love a sensible explanation. Perhaps in your package_config files you install with 'aptitude -r' instead of simple aptitude? Andreas
Re: Ubuntu hardy (8.04) problems
How does your pxe-file created by 'fai-chboot -IvF demohost' look like exactly? What kernel are you using? Andreas Olivier Le Thanh Duong writes: There was a similar problem in Gutsy, the UnionFS version shipped in the kernel doesn't support write over NFS. I got around the problem by creating the NFSROOT (and therefore the boot CD) under Feisty but putting the Gutsy installation base in it, but it's not a clean way to do it. My method is explained in more details here : http://staz.be/wiki/FAI_GUTSY (sorry I promised to migrate the HowTo to the FAI Wiki but still hasn't done so) Olivier 2008/2/6, Andreas Sindermann [EMAIL PROTECTED]: It seems that the unionfs delivered with hardy does not allow the client to change/append to existing files at all whereas creating new files is no problem... Andreas