Hey Curt,

Curt Sampson wrote:

> So we've got some Supermicro 1U boxen that don't seem to get along with
> NetBSD 1.6.1, disk-wise. Basically, when you're doing a fair amount of
> disk I/O, you start to get errors reading the disks. This has happened
> on three different machines with four different pairs of drives (the
> pairs are mirrored using RAIDFrame) and cables, configured both as one
> master drive on each IDE channel, and as a master/slave pair on one
> channel. Does anybody have any thoughts on what might be going wrong,
> and how to fix it?

I see something similar, but not quite the same.  As far as I can
tell, I haven't had any actual data loss yet - just an annoying delay
while it times out with "pciide0:1:0: lost interrupt" messages.  See
http://mail-index.netbsd.org/current-users/2003/07/07/0003.html

In summary, this is on a SuperMicro X5DAE with 82801DB IDE Controller
and a pair of IC35L120AVV207 disks (so same vendor, slightly different
controller, larger model disks in same series).
 
Simon.
--
Simon Burge                            <simonb@wasabisystems.com>
NetBSD Support and Service:         http://www.wasabisystems.com/