Charles M. Hannum --> netbsd-help (2005-01-11 18:09:21 +0000):
> On Tuesday 11 January 2005 15:07, Jukka Salmi wrote:
> >...
> > After a while, the kernel decides to downgrade wd1:
> >
> > wd1a: error reading fsbn 36275744 of 36275744-36275775 (wd1 bn 36275807; cn
> > 35987 tn 14 sn 29), retrying wd1: (aborted command, interface CRC error)
> > wd1: transfer error, downgrading to Ultra-DMA mode 5
> > wd1(viaide0:1:0): using PIO mode 4, Ultra-DMA mode 5 (Ultra/100) (using
> > DMA) cd1(viaide0:1:1): using PIO mode 4, Ultra-DMA mode 2 (Ultra/33) (using
> > DMA)
> >
> > This is quite reproducible so far.
> >
> > Hmm, does this look like a bad disk? Hints are appreciated!
> 
> It's more likely:
> 
> * a marginal cable,
> * a marginal host interface, or
> * interference from another device.
> 
> The good news is that the system is correcting for the problem and functioning 
> normally.  :-)

True ;-)

However, I'm a little concerned about what atactl prints as the SMART
error log (probably because I don't understand it): no errors for wd0,
but some for wd1 (output posted yesterday[1]; in the meantime the device
error count increased to 21). And wd1 was the drive which got downgraded
to UDMA5 by the kernel...


Regards, Jukka

[1] http://mail-index.netbsd.org/netbsd-help/2005/01/11/0003.html

-- 
bashian roulette:
$ ((RANDOM%6)) || rm -rf ~