Product guide

40
Symmetrix Fibre Channel with Qlogic HBAs on UnixWare 7
Fabric Configurations
Troubleshooting Setup Checklist step 13
This section describes some potential problems and the suggested
solutions.
Problem: I am having trouble with target 0 disk 0 of the Symmetrix.
SCO calculates the cylinder count for the first disk added to the
system improperly. SCO’s calculations for the cylinder counts must
be used. When using the disksetup command in a script, use the
cylinder count displayed in the fdisk. The EMC inq and fdisk sizes
will be slightly off for the first disk. All other disks added to the
system will have proper cylinder counts.
Problem: I am getting SCSI bus select failed console messages during normal I/O.
Example:
Apr 12 16:08:58 hostname unix: Completion code indicates "The SCSI bus
select failed"
If this condition occurs, make sure the parameter Execution Throttle
in the HBA BIOS is set to 20.
Problem: I am having trouble using the diskadd command with a large configuration.
If you are unable to perform diskadd with a large number of drives,
try running
/sbin/sdimkdtab -fi to recreate the device table
entries, or modify the kernel parameter to raise the limit for the
number of open files.
Problem: I am having trouble using my GUI interface during heavy I/O.
The SCO GUI uses a lot of memory. Increasing RAM size might help
this condition.
Problem: My host’s boot time gets progressively longer each time I reboot.
SCO does not always remove disk entries from the resource manager
after disk configuration changes. Running the following script will
force the resource manager disk table to be rebuild on reboot.
#
keys=‘/sbin/resmgr |grep vtoc |awk ’{print $1}’‘
for key in $keys; do
/sbin/resmgr -r -k $key
done
/etc/conf/bin/idconfupdate -f