Hello,

last weekend I bit the bullet and moved the last disk to my new file server.

The server is a Sparc-10/41 with 96 MB RAM, two FSBE/S, one bwtwo, and one
qfe (which I hoped would be recognized by the hme driver since AFAIK it's
simply four hme's on a single card but wasn't). It runs NetBSD-1.5.1.

The disk is an IBM DNES-318350N which has worked flawlessly for about two
years connected to an FSBE/S in a Sparc-2 running OpenBSD-2.3. It has a
single "d" partition spanning the whole disk. It is mounted in a single
unit SCSI box (type 411). It used to share the SCSI bus with a HP C2249.

Now I connected it to an FSBE/S in the Sparc-10. It shares the bus with
an IBM DDRS-39130 and an IBM DORS-32160. And - it doesn't work. Sort of.

Since disklabel recognized OpenBSD's label correctly, I let the the label
alone, and only newfs'ed it. That worked all right. But whenever I try to
actually _use_ it, for example, do fcsk or even copy some files to it, I
get an esp timeout and the disk sort of locks up.

I have changed some cables and the terminator, and since newfs succeeds,
I am inclined to think that it is not a hardware problem.

Has anybody out there seen something like this? If so, is there a fix or
a workaround?

Otherwise I'd have the choice of
a) moving that file system back to the OpenBSD box
   (hell I got that Sparc-10 to take over from two Sparc-2 server to
    make things easier - can you spell "crossmount hell")
b) try to set up a concatenated fs of 2 x 9 GB
   (I'm running out of SCSI ID's and type 411 boxes)
c) try to find another 18 GB 50 pin SCSI disk, maybe a Fujitsu
   (nearly impossible - and who knows whether it'll work)
None of these are really appealing...

	Volker