Hi,

I have pulled a SparcStation 10/30 from the dump some days ago. It came
with two Seagate "Hawk" 2G disks and Solaris 2.6, and hummed along quite
nicely for a few days.

Now, I thought about moving my two Quantum Atlas III disks with 1.5ALPHA
from the SS2 to the SS10. I got a GENERIC kernel from the latest snapshot,
test-booted it on the SS2 (just in case...) and moved the disks over.

And then I got:

scsibus0: waiting 2 seconds for devices to settle...
probe(esp0:0:0): max sync rate 10.00MB/s
sd0 at scsibus0 target 0 lun 0: <QUANTUM, QM34500TD-S, N1B0> SCSI2 0/direct
fixe
d
sd0: 4341 MB, 8057 cyl, 5 head, 220 sec, 512 bytes/sect x 8891624 sectors
probe(esp0:1:0): max sync rate 10.00MB/s
sd1 at scsibus0 target 1 lun 0: <QUANTUM, QM34500TD-S, N1B0> SCSI2 0/direct
fixe
d
sd1: 4341 MB, 8057 cyl, 5 head, 220 sec, 512 bytes/sect x 8891624 sectors

[...]

Starting syslogd.
Checking for core dump...
sd0(esp0:0:0):  Check Condition on CDB: 0x2a 00 00 75 36 34 00 00 10 00
    SENSE KEY:  Aborted Command
   INFO FIELD:  7681588
     ASC/ASCQ:  SCSI Parity Error

sd0(esp0:0:0):  Check Condition on CDB: 0x2a 00 00 75 36 34 00 00 10 00
    SENSE KEY:  Aborted Command
   INFO FIELD:  7681588
     ASC/ASCQ:  SCSI Parity Error

sd0(esp0:0:0):  Check Condition on CDB: 0x2a 00 00 75 36 34 00 00 10 00
    SENSE KEY:  Aborted Command
   INFO FIELD:  7681588
     ASC/ASCQ:  SCSI Parity Error

- later, sd1 joins in to an endless stream of parity errors.

Again, the Atlases have worked fine in the SS2 for more than a year. They
do not even have a jumper to toggle parity checking. On the other hand, the
SS10 has run just fine with its Hawks (a bit noisy, though).

I can boot single user, even fsck the partitions in peace, but as soon as I
boot multi-user (or even mount disks rw) things get noisy.

Checking the NetBSD list archives gave me nothing.

Any ideas?

	hauke

--
"It's never straight up and down"     (DEVO)