On Thu, Nov 30, 2006 at 11:04:00AM +0200, Martti Kuparinen wrote:
> Manuel Bouyer wrote:
> 
> >> Is there any possibility to get information about an harddisk error? E.g.
> >> one hardisk fails. How we get this error and can send an message to replace
> >> the broken harddisk.
> > 
> > You can use the in-tree amrctl. I have added this to my /etc/daily.local:
> > /sbin/amrctl stat -f /dev/amr0 -b
> > /sbin/amrctl stat -f /dev/amr0 -l0
> 
> I'm using this: http://mail-index.netbsd.org/current-users/2006/07/31/0007.html

Out of curiosity, how recent does the hardware need to be? I get:

# amrctl stat -g
ioctl enquiry: Input/output error

The card is pretty old:

amr0 at pci1 dev 10 function 1: AMI RAID <Series 466>
amr0: interrupting at ioapic1 pin 8 (irq 11)
amr0: firmware <3.01>, BIOS <1.36>, 16MB RAM
ld0 at amr0 unit 0: RAID 5, optimal

Should amr(4) or amrctl(8) document which are supported?

-- 
stix