I agree... it must be fixed it in systemconfigurator, even if sometimes I use the same post-install script workaround.
Have you tried to use the systemconfigurator release shipped with OSCAR? The last stable release is available here: http://svn.oscar.openclustergroup.org/oscar/branches/branch-5-0/packages/sis/SRPMS/ Otherwise you should also try with the release in the trunk: http://svn.systemimager.org/listing.php?repname=systemconfigurator&path=%2Ftrunk%2F&rev=0&sc=0 Follow the instruction to download it here: http://svn.systemimager.org/ Regards, -Andrea Patrick Hartling wrote: > I have been tinkering with SystemImager 3.7.6 and System Configurator 2.2.2, > and I have things working just about the way that I need except for the very > last step. For some reason, the GRUB boot blocks are not being written to > the client, and this of course leaves me with a system that does not boot > without some additional tweaking. An attempt is made during the > auto-installation to run grub-install, but it fails without giving me enough > information to diagnose the problem. If I run 'grub-install /dev/hda' in a > post-install script, it works fine, but it seems like it should not be > necessary to run grub-install twice. > > The output from the first run of grub-install (done by System Configurator, > right?) is as follows: > > mktemp: cannot create temp file /tmp/grub-install.log.C25188": No such file > or directory > /sbin/grub-install: line 362: $log_file: ambiguous redirect > sed: can't read /boot/grub/device.map: No such file or directory > No suitable drive was found in the generated device map. > Reverting to backed up copy. > Probing devices to guess BIOS drives. This may take a long time. > end_request: I/O error, dev fd0, sector 0 > readline() on closed filehandle IN at /usr/lib/systemconfig/Boot/Grub.pm > line 189. > Use of uninitialized value in concatenation (.) or string at > /usr/lib/systemconfig/Boot/Grub.pm line 160. > Use of uninitialized value in substitution (s///) or string at > /usr/lib/systemconfig/Boot/Grub.pm line 164. > Use of uninitialized value in concatenation (.) or string at > /usr/lib/systemconfig/Boot/Grub.pm line 235. > Use of uninitialized value in concatenation (.) or string at > /usr/lib/systemconfig/Boot/Grub.pm line 237. > Probing devices to guess BIOS drives. This may take a long time. > end_request: I/O error, dev fd0, sector 0 > > There is nothing in /tmp/si.log about GRUB, so this is the most detail as I > have been able to find about what is happening when grub-install is run by > System Configurator. > > Is it possible that grub-install is not being run from within a chroot'd > environment? The problems with accessing /tmp and /boot/grub/device.map > (which I do have) seem strange to me. At the same time, the references to > /usr/lib/systemconfig/Boot/Grub.pm suggest to me that the execution > environment should be correct, but the line numbers in the warnings don't > match up with code that would produce those warnings. How does the > auto-installation get the System Configurator version that it uses to run > grub-install? > > If it makes any difference, I am testing SystemImager within a virtual > machine. I can make the fd0 access errors go away by simply disabling the > use of a floppy disk, but that does not fix the grub-install problems. > > -Patrick ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Sisuite-users mailing list Sisuite-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sisuite-users
