Re: raidframe: Initiating parity re-write time

[Manuel Bouyer <bouyer%antioche.eu.org@localhost>]

On Tue, May 24, 2011 at 06:03:33PM +0000, Jonathan A. Kollasch wrote:
> On Tue, May 24, 2011 at 05:49:53PM +0000, Michael van Elst wrote:
> > uhel%gmx.net@localhost writes:
> > 
> > >The time of initiating the re-write of parity with 
> > >'raidctl -v -i raid0' took ~64 hours. 
> > >Is that a timeframe one could expect?
> > >The system is an Intel atom and the size of the disks is 698GB.
> > 
> > You should expect something like 5 hours.
> > 
> > >pciide0: bus-master DMA support present, but unused (no driver support)
> >                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> > Doesn't look nice and might be the reason for the slowness.
> 
> Probably is, that'll force you back to PIO.
> 
> > 
> > >pciide0: vendor 0x10de product 0x0ab4 (rev. 0xb1)
> > 
> > That's NVIDIA chipset and apparently not fully supported even in -current.
> 
> I see some indication it's AHCI-compatible.  Probably needs IDs added to
> ahcisata_pci.c.

Or maybe there's a bios option to turn it to ahci mode ?

-- 
Manuel Bouyer <bouyer%antioche.eu.org@localhost>
     NetBSD: 26 ans d'experience feront toujours la difference
--