I was given a machine with six scsi drives and I wanted to play around with RAID setups. The system as given to me was setup with 5 drives in a raid and one used as /boot. I wanted to use all of them for the raid, so I added a large ATA drive. On the new drive I made a /boot partition, a swap partition, a partition to use as a spare in the RAID and the rest of the space (90% of the drive) as /home.
I setup the raid and everything else (with stable) and it seemed good, on reboot i discovered a problem: that nice new large drive was too large for the mainboard's BIOS to recognize and so the /boot partition could not boot! I thought positively, ooh a learning opportunity. I found a small drive the system could recognize and created a filesystem on it, copied over all the files from the unbootable /boot partition, using rescue mode. I assembled the RAID drive, and ran a shell on the resulting /dev/md0. Modified the /etc/fstab so /boot would have the correct blkid. I then ran grub-install on the small drive (/dev/sdg). During the grub-install it spewed 25 errors "error: unknown LVM metadata header". But the command continued on and ended saying "Installation finished. No Errors reported." After rebooting, it sees the small drive and grub is started, but it pauses a moment says "Welcome to grub" then just dumps me into a grub shell. Using the boot command I get "error: no loadable kernel". Do I have the right idea on fixing this /boot problem in general? Or did I just fail to do some minor thing? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/558313664.275.1389966436314.JavaMail.sas1@[172.29.241.245]
