Bug#568975: ([squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable)

2010-02-17 Thread Frans Pop
On Thursday 11 February 2010, Frans Pop wrote: > On Thursday 11 February 2010, vincent.mcint...@csiro.au wrote: > > Should this be reassigned to partman-auto-raid? > > I would say partman-md (as it will also affect manually created arrays). > As p-a-raid depends on partman-md that will work fine. >

Processed: Re: Bug#568975: ([squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable)

2010-02-11 Thread Debian Bug Tracking System
Processing commands for cont...@bugs.debian.org: > reassign 568975 partman-md 49 Bug #568975 [installation-reports] [squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable Bug reassigned from package 'installation-reports' to 'partman-md'. Bug #568975 [partman-md] [squeeze/lenny

Bug#568975: ([squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable)

2010-02-11 Thread Frans Pop
reassign 568975 partman-md 49 thanks On Thursday 11 February 2010, vincent.mcint...@csiro.au wrote: > Should this be reassigned to partman-auto-raid? I would say partman-md (as it will also affect manually created arrays). As p-a-raid depends on partman-md that will work fine. We'll possibly nee

Bug#568975: ([squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable)

2010-02-10 Thread Vincent . McIntyre
On Thu, 11 Feb 2010, vincent.mcint...@csiro.au wrote: The options for resolving it (so lenny-via-squeeze works) seem to be: - add a debconf configuration item to specify the metadata version I started a patch but I have absolutely no idea what I'm doing. The attached is just an attempt to cla

Bug#568975: ([squeeze/lenny][i386][2010-02-04] grub-installer fails, system unbootable)

2010-02-10 Thread Vincent . McIntyre
I think I've identified the issue here. Details at the end. In short: partman-auto-raid v15 creates /dev/md0 with a v1.1 superblock. The version of grub being used (0.97-47lenny2) doesn't handle this version of the superblock metadata AFAICT (see e.g. #554500). This seems to be causing the gru