Bug#583917: Bug#594472: grub-pc: scary messages and very long boot time after upgrade

[Agustin Martin]

2010/8/26 Toni Mueller <support at oeko.net>:
> Package: grub-pc
> Version: 1.98+20100804-2
> Severity: important
>
>
> Hi,
>
> my machine is a standard PC with two SATA disks, bundled as md0 and md1,
> with /boot on md0, and everything else in a LUKS encrypted container on
> md1 with LVM inside. So, booting normally works by fetching the kernel
> and initrd from /boot, then asking for a passphrase to unlock the
> container, then proceed with the boot process.
>
> Yesterday, I upgraded to all the latest and greatest in Testing, which
> included both new versions of grub-pc and
> linux-image--2.6.32-5-686-bigmem (I have 8GB RAM in my machine) and
> grub-pc presented me with a statement like "your UUIDs are garbled. I
> need to regenerate them, or you'll be unable to boot.", with my two
> disks selected as the default devices to regenerate UUIDs for. I
> accepted, understanding that grub-pc wanted to write new MBRs to the
> disks, then shut down the machine after the upgrade finished. Today, the
> machine booted to the point where it would normally ask for the
> passphrase, but instead, it said:
>
>  Volume group "ev0" not found
>  Skipping volume group ev0
> Unable to find LVM volume ev0/root
>  Volume group "ev0" not found
>  Skipping volume group ev0
> Unable to find LVM volume ev0/swap
>
> After several (5?) minutes where the machine didn't do anything I was
> able to see, the screen was suddenly flushed with a large number of
> messages like:
>
>  /sys/devices/virtual/block/md0 (10715)
>  /sys/devices/virtual/block/md1 (10716)
>  /sys/devices/virtual/block/md0 (10717)
>  /sys/devices/virtual/block/md1 (10718)
>  /sys/devices/virtual/block/md0 (10719)
>  /sys/devices/virtual/block/md1 (10720)
>  /sys/devices/virtual/block/md0 (10721)
> Unlocking the disk /dev/md1 (md1_crypt)
> Enter the passphrase:
>
> After entering the passphrase, the machine used several more minutes
> while the cpu cooler went into high pitch, then it emitted the following
> message:
>
>  10 logical volume(s) in volume group "ev0" now active
> cryptsetup: md1_crypt set up successfully
>
> Then it took some more time, after which it emitted a flurry of
> messages, ending with this:
>
>
>  /sys/devices/virtual/block/md0 (25598)
>  /sys/devices/virtual/block/md1 (25599)
>  /sys/devices/virtual/block/md0 (25600)
>  /sys/devices/virtual/block/md1 (25601)

May this be related to #583917,  mdadm: long delay (6-200 minutes)
during boot (root device detection) after upgrade on RAID/LVM/LUKS
setup?

Does downgrading to mdadm 3.1.1 help?

-- 
Agustin