Installation guide

Sun StorageTek Common Array Manager Software Release Notes 33
Cache Stop % Cannot be Greater than Cache Start %
Bug 6590637 –Attempting to modify the Cache Start% and Cache Stop% parameters
from the array’s Administration page so that the value assigned to Cache Stop% is
greater than the value assigned to Cache Start% results in the error message
“setCacheParams operation failed:43”.
Workaround – Use valid values. Since the Cache Stop% is the percentage of
unwritten data in the cache that will stop a cache flush that is currently in progress,
it must not be greater than the value for Cache Start%, which is the percentage of
unwritten data in the cache that will trigger a cache flush.
Cannot Cancel Some Jobs
Bug 6600387 –Some jobs, like volume creation jobs, cannot be cancelled on the array
once they have started. However, if multiple jobs are queued up for an array, any job
in the queue can be cancelled before the array actually starts the job.
CLI - When Performing an Array Import, Do Not Modify Management
Objects
If you create management objects while an “import array” job is running, it might
interfere with the import. Be sure that everyone who uses the destination array does
not modify or create any objects (including volumes, initiators, mappings, and so
on) while the import is in progress.
Controllers Reboot when Collecting Support Data
Bug 6649629 –When data is collected and a drive is by-passed, one or both
controllers might reboot. This can also happen when an array (containing a drive
that was part of a virtual disk) is removed while support data is collected. When
support data is captured, satecapturedata.dmp file is also captured. This .dmp file
contains a set of shell commands that are run on the controller. The “hids 108”
command, added in release 6.0, can reboot the controller when run while a drive in
a system is by-passed, or when a drive that was part of a virtual disk is physically
removed.
Workaround – Avoid collecting support data during times of heavy use. The
controller reset will only occur upon the capture of the support data. Individual
collections of Alarms, Array Profiles, and Major Event Logs are possible within each
of the management utilities. There is no completely successful workaround at this
time and this issue remains under investigation.