HP StorageWorks 12000, Руководство пользователя

Страница: 229 / 266

Solution Possible causes Symptom
5. Repeat this procedure for each
server visible to each SDLT
tape drive.
None
This is expected behavior and does
not indicate a problem.
At reboot, there are spurious critical
FC port failures reported as notific-
ation alerts, usually on every port.
Later, Info notification alerts for
each FC host port are generated,
indicating the FC ports are operat-
ing normally.
If a blank drive is installed in drive
bay 0, the system will not boot. To
remedy this, switch the drive loc-
ated in bay 0 with the drive located
in bay 1.
A blank drive was installed in drive
bay 0. After replacing drive 0 in a node,
the system will not boot.
Re-enable oversubscription, set the
alert threshold so that the storage
pool is within the capacity
threshold, then wait for the pool to
return to good (green) status. Dis-
able oversubscription and the stor-
age pool will maintain good status.
When you disable oversubscrip-
tion, the system does not monitor
the storage pool's capacity, even
to check that it is no longer at a
critical level. The status will not
change.
When oversubscription is enabled
and you create enough cartridges
in a storage pool to put your free
storage capacity below the
threshold, the pool may reach crit-
ical status. If you then disable
oversubscription, your storage pool
will remain in critical status and will
not return to good status—even
after reboot.
Automigration/Replication Issues
Although Automigration is configured and managed through the Automigration/Replication tab on
Command View VLS, errors and events are reported through the usual notification alerts. See
Notification Alerts .
IMPORTANT:
If a destination library is directly connected to the VLS (
LOOP
mode), and you disconnect the destination
library to change its connection to be via a SAN (
FABRIC
mode), you will need to reboot the VLS in
order for this change to work. If you do not reboot, the destination library will be marked as
FAILED
after you disconnect and reconnect it.
NOTE:
The “mirror_broken” state can occur when there is not enough room on the virtual tape to create the
copy, there are no available slots to create the copy, the library cannot read the header of the physical
tape, the library cannot determine that a tape without a header is blank, or the library cannot
successfully write the header to the virtual tape.
HP StorageWorks 229