[CentOS-docs] I'd like to contribute two wiki articles

Mon May 4 12:36:04 UTC 2009
Lev Shamardin <shamardin at gmail.com>


I've finished the first version of the article here:


On 05/04/2009 01:35 PM, Ralph Angenendt wrote:
> > Lev Shamardin wrote:
>> >> On 04/29/2009 10:55 PM, Ralph Angenendt wrote:
>>> >>> Is that different from
>> >> There is one really HUGE difference. Using that guide you get a system with N
>> >> regular raids (where N is the number of the partitions you create). This means
>> >> that to rebuild a raid after failure you need to recreate N partitions on a
>> >> spare disk, and rebuild N raid volumes.
>> >>
>> >> My guide describes how to install on a partitionable mdadm raid (with devices
>> >> /dev/md_d0pX). To repair such system you need to repair only one raid volume,
>> >> and you do not need to mess with partition table, etc.
> >
> > Hmmm. I guess those are better done as two separate HowTos? Sorry for
> > the delay, we had a holiday over here and I was mostly away from the
> > computer over the weekend.

I think it is better to leave these as two separate HowTos because they describe
two different setups. The traditional one is well supported by default CentOS,
but requires more effort in case of failures. The partitionable raid is more
fault tolerant but requires patching mkinitrd (check

> > Maybewe can merge the two documents somehow, but before thinking about
> > that, I need to know what is in the article  :)

And what about the second guide? (about repairing failed raid5 with more then 1
failed drive)

> > Ah yes, Hope you are fine with the Creative Commons license.

That's just fine.