Re: long raidctl initialization times

[Sean Hafeez <sahafeez@zaphodb.org>]

I did init'd a 4x160 ATA133 RAID5 in about 4 hours.

-Sean





"Fiat justitia ruat caelum."


On Aug 13, 2004, at 12:48 AM, Aaron Glenn wrote:

> All,
>
> What would cause a 16+ hour array initialization?
>
> I've built 74GB mirrored arrays three times on identical hardware with
> 3.5-RELEASE and each time the intialization or rebuilding has taken
> anywhere from 10 to 16 hours to complete. I don't have a dmesg handy;
> however the two serial ATA drives are attached to an Intel
> 6300ESB-based on board controller which comes up "unidentified". I
> checked cvs and noticed the PCI ID's for the 6300ESB controller were
> added to pcidevs.h a week ago.
>
> Is it possible to update my pcidevs.h in 3.5 source tree and recompile
> my kernel? Is the fact my controller isn't being recognized causing
> the insane initialization times?
>
> In order to get OpenBSD installed on the hard drives I had to "Map
> SATA to 'Legacy PATA'", otherwise the installer could not find the
> hard drives. While I'm pretty sure this is the cause of my problem,
> I'm certainly not certain.
>
> Regards,
> aaron.glenn