[Gluster-users] XFS and MD RAID

[Brian Candler]

On Mon, Sep 10, 2012 at 11:03:41AM +0200, Stephan von Krawczynski wrote:
> Brian, please re-think this. What you call a stable kernel (Ubuntu 3.2.0-30)
> is indeed very old.

I am talking about the official kernel for the Ubuntu 12.04 Long Term
Support server release.

If you're saying that Ubuntu chose a dog to base their LTS release from,
then that's unfortunate.

The fact that they back-ported some "fix" from a later kernel which broke
hot-swap is also unfortunate.  Does this mean that Ubuntu should do more
testing themselves, or is it reasonable for them to trust the upstream
kernel team to test their work properly?  That's an open question.  However
in this case, it seems to me that *nobody* tested it properly.

Why should I believe that work committed to 3.4.x is tested properly, but
work commited to 3.2.x is not - especially when such work has probably been
committed to 3.4.x first and then back-ported?

> I must say I would probably drop them only because current processors are
> faster with MD anyway. I just built a box with XEON E3-1280v2 with MD raid
> 4x2TB and I am impressed by the performance.

I'm sure that performance of MD RAID is easily as good as, if not better
than, a RAID controller card with some old CPU on it.

Please understand that I *do* really want to use MD RAID. It's a doddle to
configure and monitor compared to all those proprietary admin and monitoring
tools out there, and it gives a consistent interface across different
machines.  But RAID exists only to protect against failures, and it is
worthless if the failure cases are not handled properly.

Regards,

Brian.