F11 non-LVM /boot partition not accessible

[marcus hall]

I have a system updated from F9 -> F10 -> F11 that no longer can get access
to the /boot partition (the F11 update was the trigger for this).  After the
update, on the reboot, grub saw the /boot partition just fine and booted
the kernel successfully, then when the kernel went to check and mount the
filesystem, it couldn't find it.  I commented out the filesystem in /etc/fstab
and things came up OK, but of course I will have difficulty applying any
new kernel images in the future..

The system has two mirrored 320GB SATA drives.  They are partitioned
identically, with a ~200MB partition 1 for /boot and the rest of the disk
allocated for LVM:

Disk /dev/sda: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000007b4

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          25      200781   83  Linux
/dev/sda2              26       38913   312367860   8e  Linux LVM


fdisk is happy with the partition table on both disks.  During boot, the
kernel seems to recognize both partitions:

scsi 0:0:0:0: Direct-Access     ATA      WDC WD3200YS-01P 21.0 PQ: 0 ANSI: 5
sd 0:0:0:0: Attached scsi generic sg0 type 0
sd 0:0:0:0: [sda] 625142448 512-byte hardware sectors: (320 GB/298 GiB)
sd 0:0:0:0: [sda] Write Protect is off
sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
 sda: sda1 sda2
sd 0:0:0:0: [sda] Attached SCSI disk
ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
ata2.00: ATA-7: WDC WD3200YS-01PGB0, 21.00M21, max UDMA/133
ata2.00: 625142448 sectors, multi 0: LBA48 NCQ (depth 1)
ata2.00: configured for UDMA/133
scsi 1:0:0:0: Direct-Access     ATA      WDC WD3200YS-01P 21.0 PQ: 0 ANSI: 5
sd 1:0:0:0: [sdb] 625142448 512-byte hardware sectors: (320 GB/298 GiB)
sd 1:0:0:0: Attached scsi generic sg1 type 0
sd 1:0:0:0: [sdb] Write Protect is off
sd 1:0:0:0: [sdb] Mode Sense: 00 3a 00 00
sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
 sdb: sdb1 sdb2

By the time user space gets started, however, the kernel and/or udev have
created only /dev/sda, /dev/sdb, and /dev/sdb2.  If I manually create a
/dev/sda1 or /dev/sdb1, they cannot access the partitions on the disk and
return an immediate error ENODEV.

I have other systems running just fine that have followed the same update
path (or even longer), but this is the only system I have that has mirrored
disks or LVM.  Has there been any change that might have caused this?  I am
a little suspicous that something is aware of the disk mirroring, since there
is a sdb2 device file but no sda2 file.  The LVM table is based on /dev/sdb2,
not on /dev/sdb itself.  Is this the normal way, or is there some way that
this could be causing the trouble?

Any suggestions of what to look at, or ideas what might be the culprit here?

Thanks in advance!

Marcus Hall
marcus at tuells.org