The MD roadmap

By Jonathan Corbet
February 16, 2011

Users of the MD (multiple disk or RAID) subsystem in Linux may be interested in the MD roadmap posted by maintainer Neil Brown. It discusses a number of things he has planned for MD in quite a bit of detail; as Neil put it:

A particular need I am finding for this road map is to make explicit the required ordering and interdependence of certain tasks. Hopefully that will make it easier to address them in an appropriate order, and mean that I waste less time saying "this is too hard, I might go read some email instead".

There are a lot of enhancements in the pipeline. A bad block log would allow RAID arrays to continue functioning in the presence of bad blocks without needing to immediately eject the offending drive. There is a variant on "hot replace" which would allow a new drive to be inserted before removing the old one, thus allowing the array to continue with a full complement of drives while the new one is being populated. Tracking of areas which are known not to contain useful data would reduce synchronization costs. A number of proposed enhancements to the "reshape" functionality would make it more robust and flexible and allow operations to be undone. A number of other changes are contemplated as well; see Neil's post for the full list.

Index entries for this article
Kernel	Block layer/RAID
Kernel	RAID

to post comments

The MD roadmap

Posted Feb 21, 2011 22:15 UTC (Mon) by PhracturedBlue (subscriber, #4193) [Link] (1 responses)

I thought there was an effort to merge the 'md-raid' capabilities into the device-mapper system so that it would be possible to manage raid and virtual-partitioning through LVM.
I searched and couldn't find the status of that implementation, and am not sure if it is independent of Neil's work or if the change already happened/is not going to happen and Neil's work is on the 'one true software raid' driver.

I am definitely interested in the hot-replace capability. Each time I upgrade the size of my raid, i get scared at having the array downgraded for so long while recomputing all data.

The MD roadmap

Posted Feb 22, 2011 0:08 UTC (Tue) by neilbrown (subscriber, #359) [Link]

The status of that effort is that there will be a 'raid' target available to dm devices in 2.6.38 which will provide access the md raid5 implementation.

It is likely that this will be extended to cover md raid1 as well in due course.

As features are added to md-raid they would then become available to dm-raid once the dm-raid configuration interface is enhanced to understand them.

i.e. once hot-replace is implemented for md, all that would be required for it to be available under dm would be a mechanism in the dm interface to request it.

The MD roadmap

Posted Mar 9, 2011 17:37 UTC (Wed) by Terra (guest, #73487) [Link] (1 responses)

Does anyone know when the Linux kernel will support RAID/MD version 1.x superblock autodetecting during bootup?

Ref: https://raid.wiki.kernel.org/index.php/RAID_superblock_fo...

The new mdadm tools default to version 1.x, and we found our newly deployed servers unable to boot, so we got bit by this issue the hard way and had to go back and recreate them with superblock version 0.90

The MD roadmap

Posted Mar 9, 2011 17:52 UTC (Wed) by nix (subscriber, #2304) [Link]

As far as I know, there are no plans ever to implement this, though Neil is not averse to someone else doing the work. But you don't need it: it took me three hours to come up with an initramfs that could boot my RAID arrays, and after that I didn't care what the kernel's built-in /-mounting code did. (Virtually every distro already has initramfses / initrds that do this.)