| 1 |
On Wed, 14 Mar 2007 17:52:26 -0700 |
| 2 |
kashani <kashani-list@××××××××.net> wrote: |
| 3 |
|
| 4 |
> Dan Farrell wrote: |
| 5 |
> |
| 6 |
> > in reality, though, I think the best performance would probaby |
| 7 |
> > involve just using the fast drive. RAID introduces too much |
| 8 |
> > overhead to make up for itself in this situation I think. |
| 9 |
|
| 10 |
> I'm betting the act of seeking across the platters on the fast drive |
| 11 |
> for two separate partitions on the disk makes performance truly |
| 12 |
> awful. The idea of separate stripe sizes in the original post makes |
| 13 |
> sense for shifting more I/O to drive A, the fast one, without causing |
| 14 |
> weird things to happen on the physical disc that the software can't |
| 15 |
> optimize for. |
| 16 |
I wouldn't mind seeing test results, but that argument has some weight |
| 17 |
to it. |
| 18 |
> However I think messing with stripe sizes is not something Linux |
| 19 |
> software raid (or any hardware raid I've dealt with) supports. |
| 20 |
This is a nice idea, but I haven't been able to find evidence of |
| 21 |
supporting it. |
| 22 |
> kashani |
| 23 |
-- |
| 24 |
gentoo-user@g.o mailing list |
Archives