6760 ServerNet/DA Manual

Operating the 6760 ServerNet/DA
6760 ServerNet/DA Manual—424879-003
4-19
Power-Failure Recovery
for Himalaya S-Series Servers, and the manufacturer’s documentation shipped with
the product.
Power-Failure Recovery
NonStop S-series systems have the following power-failure protection: The server,
which includes all components in the processor enclosures and I/O enclosures, can
stay operational for up to 30 seconds during a power outage. The actual time depends
on the condition of the batteries in the enclosure. For detailed information about this
feature, see NonStop S-Series Planning and Configuration Guide.
This means that during any short power outage, the 6760 adapter remains operational
while the modular disk and tape subsystems that were migrated from the NonStop
K-series systems do not. This situation could result in data integrity problems if the
system software continues processing data from a disk drive or tape drive during a
short power outage.
To avoid any possibility of a data integrity problem, any time that the system software
detects a power outage, the device state for the external disk drives and tape drives is
changed to DOWN. When power is normal again, the operator must use SCF
commands to restart these devices to bring them back online.
If you want an alternative to this manual procedure, you must provide an
uninterruptible power supply (UPS) for the modular disk and tape subsystems. To use
this alternative, you must also configure the storage subsystem by using an SCF
command that tells the system software that this capability exists. For more
information about the UPS attribute of the ALTER SUBSYS command, see the SCF
Reference Manual for the Storage Subsystem.
Troubleshooting Adapter and Device Problems
The purpose of this subsection is to help you determine which component in the
storage subsystem is failing. To do that, you use the TSM package and the SCF
STATUS and SCF INFO commands.
Preparing to Test a 6760 Adapter and Its Devices
Before you test a 6760 adapter and its attached devices, make sure that the following
processes are running:
•
$ZTSMS (SNMP process)
•
$ZTSM (TSM process)
•
$ZZSTO (storage subsystem manager process)