NonStop Operations Guide for H-Series and J-Series RVUs

The processor in the FROM row is down or nonexistent. For a processor that does exist on

your system, this status is abnormal.

• ERROR nnn (for an entire row)

The processor in the FROM row unexpectedly returned a file-system error to that ServerNet

fabric.

• UNA (unavailable)

The path from the processor in the FROM row to the processor in the TO column is down

because the processor in the TO column is down. For a processor that does exist on your

system, this status is abnormal. UNA overrides all other states.

Recovery Operations for the ServerNet Fabrics

For most recovery operations, refer to the SCF Reference Manual for the Kernel Subsystem.

Recovery Operations for a Down Disk Due to a Fabric Failure

When a path to a disk drive goes down due to a ServerNet fabric failure (either the ServerNet X

or Y fabric is down), the storage subsystem automatically switches the paths of the disk drive if

possible, so that the disk drive remains operational. This switching might result in a disk drive being

placed in the STOPPED state with a substate of HARDDOWN.

You must restart any disk path that was using the fabric that went down. Otherwise the storage

subsystem never attempts to use that path, which creates a potential single point of failure. For

more information, refer to “Recovery Operations for a Down Disk or Down Disk Path” (page 152).

Recovery Operations for a Down Path Between Processors

When the status is either DIS (disabled) or DN (down), you can restart all paths between processors

on the X fabric or Y fabric:

> SCF START SERVERNET $ZSNET.{X|Y}.*

Refer to the SCF Reference Manual for the Kernel Subsystem.

Recovery Operations for a Down Processor

If the status for an existing processor is <- DOWN or UNA, refer to “Recovery Operations for a

Processor Halt” (page 130).

Recovery Operations for a File-System Error

For information about file-system errors, refer to the Guardian Procedure Errors and Messages

Manual.