This is a metabug for other bugs I filed against grub-installer,
partman-auto, partman-base, mdadm and lvm2.

I noticed that linux mdraid has working support for Intel Matrix Raid
(aka Intel Rapid Storage Technology), called "imsm" in mdadm. This
is a form of fakeraid, like dmraid.

Dmraid has never been really supported in Debian, and upstream
development has also ceased. It looks like mdraid is the way
forward.

I have added support for installing on mdraid based Intel Matrix
raid to the debian-installer, the lvm2 package, and mdadm.

With these patches applied I can install and boot Debian Wheezy without
any trouble on an Intel Matrix RAID RAID1 array - even when
combined with lvm.

Patches are available, the bug reports will be tagged as blocking
for this metabug, and are:

grub-installer:
    - also detect mdadm fakeraid underneath lvm, if it's just one array
    - install grub on each disk of the underlying array
    - add md/0 entry to device.map so grub-probe works (for upgrades etc)

partman-auto:
   - don't hide /dev/md* arrays if they are provisioned on whole-disk
     devices - in that case the md array itself is partitionable,
     e.g. for md-fakeraid like imsm (Intel Matrix raid).

partman-base:
   - init.d/parted: Skip devices that are part of a mdraid device

lvm2:
   - lvm2 already filters mdraid superblocks, now also filter IMSM superblocks

Mike.

31.01.2013 16:14, Miquel van Smoorenburg wrote:
[]

Out of curiocity, did you actually test the latest mdadm package?
I did a few more changes on top of your patch, but I don't have any
hardware where I can try it, so I've no idea if it actually works
or not...

Thanks,

/mjt

Yes, I did. Yesterday I took the latest snapshot from the wheezy
installer media, and replaced the debs/udebs with the ones that I
updated, replaced mdadm with the latest version from unstable (-5), and
built a new image.

It all worked fine, installing, booting, rebooting, etc.

Thanks,

Mike.

Hi Mike,

in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=699430#17 you said your
udebs worked, but all four blocking bugs of this one are still open, so am I
right to assume this bug is still present in both wheezy and sid images?


cheers,
	Holger

Hi Mike,

in http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=699430#17 you said your
udebs worked, but all four blocking bugs of this one are still open, so am I
right to assume this bug is still present in both wheezy and sid images?


cheers,
	Holger

control: block -1 by 699431 699432 699434 684712 691710
thanks