md: RAID-6 patch available for testing

md: RAID-6 patch available for testing

[H. Peter Anvin]

[Already announced to linux-raid, but I thought it might get wider
distribution in this list.]

For those that don't know, I've been working on adding RAID-6 (dual
failure recovery) to the md system for a while.  It started out as a
project because the math was interesting, and Penguin Computing for
donated a very much needed test system (thanks!)

Well, at least I have a piece of code that passes my relatively simple
functionality tests.  Still, that's news, and this is the first RAID-6
snapshot that isn't *known* to be broken :)

I can at least mount filesystems, read and write data, reboot the
system and have the data still there, with 1 or 2 disks lost, and do a
reconstruction once the drives are added back in.

New development snapshot at:

ftp://ftp.kernel.org/pub/linux/kernel/people/hpa/raid6-20031224c-experimental.tar.gz

Please test it out and let me know how badly it sucks :)

At some point I'll try to run some benchmarks.  There is also a lot of
optimization still to be done.

       -hpa
-- 
<hpa@transmeta.com> at work, <hpa@zytor.com> in private!
If you send me mail in HTML format I will assume it's spam.
"Unix gives you enough rope to shoot yourself in the foot."
Architectures needed: ia64 m68k mips64 ppc ppc64 s390 s390x sh v850 x86-64

Re: md: RAID-6 patch available for testing

[Bill Davidsen]

H. Peter Anvin wrote:
> [Already announced to linux-raid, but I thought it might get wider
> distribution in this list.]
> 
> For those that don't know, I've been working on adding RAID-6 (dual
> failure recovery) to the md system for a while.  It started out as a
> project because the math was interesting, and Penguin Computing for
> donated a very much needed test system (thanks!)
> 
> Well, at least I have a piece of code that passes my relatively simple
> functionality tests.  Still, that's news, and this is the first RAID-6
> snapshot that isn't *known* to be broken :)
> 
> I can at least mount filesystems, read and write data, reboot the
> system and have the data still there, with 1 or 2 disks lost, and do a
> reconstruction once the drives are added back in.
> 
> New development snapshot at:
> 
> ftp://ftp.kernel.org/pub/linux/kernel/people/hpa/raid6-20031224c-experimental.tar.gz
> 
> Please test it out and let me know how badly it sucks :)
> 
> At some point I'll try to run some benchmarks.  There is also a lot of
> optimization still to be done.

If this works as well as you intend, unless the performance is really 
broken it will still be useful in situations where data loss would be 
very painful. It wouldn't be bad in ad-hoc arrays I build from drives 
which have been pulled because they have too many POH or power cycles;-)

-- 
bill davidsen <davidsen@tmr.com>
   CTO TMR Associates, Inc
   Doing interesting things with small computers since 1979