My source server was a cluster node running 6.0 SP5, and I'm trying to upgrade it to 6.5 SP5. Remote Upgrade failed twice during the file copy when running it direct from the CDs. So I copied the CDs to another NetWare server and ran it from there, and the file copy succeeded.

Now that the server has rebooted, it is ABENDing in HDETECT.NLM on the storage driver QL2300.HAM. The ABEND screen says the ABEND occurred in the IDLE Process.

Here's the original STARTUP.NCF that was working. The SYS volume is on an internal RAID of 3 drives accessed with IPSRAID.NLM. The cluster volumes are on a SAN, accessed through the QL2300 and QL2200 drivers. The IBMSAN.CDM takes care of the multi-path failover in lieu of Novell's MM.NLM

set multi-path support = off
LOAD MPS14.PSM
LOAD IDECD.CDM
LOAD SCSIHD.CDM
LOAD IBMSAN.CDM
LOAD IDEATA.HAM PORT=1F0 INT=E
;LOAD ADPT160M.HAM SLOT=10010
;LOAD ADPT160M.HAM SLOT=10011
LOAD IPSRAID.HAM SLOT=4
LOAD QL2300.HAM SLOT=5 /LUNS /ALLPATHS /PORTNAMES /xretry=400
LOAD QL2200.HAM SLOT=3 /LUNS /ALLPATHS /PORTNAMES /xretry=400

I tried circumventing HDETECT by commenting it out of c:\nwinst.tmp\instauto.ncf, and putting the above statements in, but the server still abends.

Another strange thing happens when stepping through the boot sequence with Maintenance mode. When NSS.NLM loads for the 1st time it complains that the SYS volume could not be mounted, and "/AutoShiftHash=17 is not available in the current NSS configuration".

Any ideas on getting this upgrade to proceed?