NonStop S-Series Operations Guide (G06.27+)

Overview of Monitoring and Recovery

HP NonStop S-Series Operations Guide—522459-008

3-25

Using SCF to Locate the Problem

The system displays:

The primary path is currently active. Because the primary path is configured for the X

fabric, you conclude that the disk first tried to use the Y fabric and failed before

switching to the X fabric. Now you want to find out why the path switch was not

transparent.

At this point, you recall that if the storage subsystem’s direct bulk I/O feature is set to

ON, a disk can bypass the disk process and perform a direct transfer of bulk I/O over

the preferred path. You execute an SCF INFO command to display attributes for the

storage subsystem:

> SCF INFO SUBSYS $ZZSTO

The system displays:

The direct bulk I/O feature (BulkIO) is ON. You conclude that the disk attempted a

direct bulk I/O transfer across its active path, which was initially configured through the

Y fabric. Without the services of the disk process, the communications failure and

subsequent path switch were not transparent, and the disk generated the 14 errors.

You decide that the bulk I/O request probably came from a BACKUP/RESTORE

process running in a processor in group 01.

From all this information about where the breakdowns occurred, you conclude that

there is a hardware failure related to the Y fabric, in either group 01 or group 03—

probably in port 2 of the SEB in slot 52.

STORAGE - Status DISK \GATE8.$D3101-*

LDev Path Status State Substate Primary Backup

PID PID

96 PRIMARY ACTIVE STARTED 4,263 5,263

96 BACKUP INACTIVE STARTED 4,263 5,263

96 MIRROR INACTIVE STARTED 4,263 5,263

96 MIRROR-BACKUP ACTIVE STARTED 4,263 5,263

1-> SCF INFO SUBSYS $ZZSTO

SCF - T9082G02 - (16JAN01) (08JAN01) - 07/09/2001 13:49:36 System \COMM

STORAGE - Info SUBSYS $ZZSTO

AutoConfigure AutoRevive Autostart BulkIO LabelTape UPS

OFF OFF ON ON ON OFF

Total Errors = 0 Total Warnings = 0