Original question:
We have a problem connecting some RZ29L 4Gb disks, mounted into a BA346 box
and connected to an Alpha AXP150. We can write onto this RZ29L, but doing
an fsck or reading the disk afterwards hangs the system:
fsck -o /dev/rz2c
/sbin/ufs_fsck -o /dev/rz2c
** /dev/rz2c
** Last Mounted on
** Phase 1 - Check Blocks and Sizes
cam_logger: CAM_ERROR packet
cam_logger: bus 0 target 2 lun 0
Ss_perform_timeout
Timeout on disconnected request
.
.
Reached max abort count, scheduled bus reset
aha_bus_reset
Resetting the SCSI bus at request
I/O on an RZ28 - also in this box mounted - show no problem at all. The
problem must be on the alpha-side, because we connected the BA-box
including cable and disks to an NT system and could read/write without any
problems.
We are running Dunix 3.2G (cannot yet upgrade). The AXP150 has the latest
firmware (2.2) loaded. According to local Digital-support this
configuration is allowed but they cannot help us on another way as swapping
HW (SCSI Ctrl, BA-Box, dsk) which didn't bring anything.
Has somebody seen this problem before or maybe even know a solution?
Any help is welcome
Summary:
The official answer from DEC is: "Engineering confirms that this
configuration won't work. The Adaptec controller does only support RZ29 up
to RZ29B. The RZ29L will not work. Proposed solution: add a newer Adaptec
controller to your AXP150/DEC2000. As in this configuration you are not
able to boot from the newer controller, you still have to use the elder
controller as well"
Thanks to
Vinzenz_Esser_at_emass.com
alan_at_nabeth.cxo.dec.com
and DEC
for their help
Hans van der Heide (hans.vanderheide_at_bluewin.ch)
BDL Informatik GmbH, Switzerland
Received on Tue Mar 24 1998 - 07:28:56 NZST