Ciprico 6500 manual Drive Failure Replacement Rebuilding, Determining Drive Failure

Chapter 4 Maintenance, Troubleshooting & Hardware Replacement

Drive Failure, Replacement, and Rebuilding Page 4 - 21

Drive Failure,

Replacement,

and Rebuilding

The following pages contain information you will need to handle 6500 array drive failures, drive replacement, and drive rebuilding.

When all 9 drives are installed in the subsystem, the failure of one drive does not affect performance. The failure is reported through the display/operation panel. An alarm will sound if enabled.

The loss of a second drive can cause a catastrophic loss of data. To ensure continued array operation without data loss, a failed drive should be replaced and its data rebuilt as quickly as possible.

Determining Drive Failure

CAUTION If you are formatting a drive or rebuilding data on a drive,

•do not reset the array by any method,

•do not remove the array's SCSI terminator, and

•do not turn off the power to the array.

The array uses the following conditions to identify a failed drive:

•The drive does not pass drive tests or fails in command time-out.

•A single drive is unformatted or from another set.

•Drive error thresholds.

Reporting Drive Failure

If any of the above conditions are met, the controller will spin-down the drive and report the failure.

In the 6500 array, drive failures create a SCSI Unit Attention condition in the array if the 537$65(&29(5option is not enabled and unit attention reporting is enabled.

Following is the conditional logic of reporting for drive failures. In the event of a drive failure: if the alarm is enabled, then the alarm will sound; if recovered error reporting is enabled, then a recovered error will be reported; if recovered error reporting is not enabled but Unit Attention reporting is enabled, then a Unit Attention condition will be reported; if neither recovered error reporting nor Unit Attention reporting are enabled, then no error will be reported.

During a Unit Attention condition, the next command received from each initiator is ignored and the array reports a Check Condition status. When the host requests sense data, it will receive an Additional Sense Code indicating the general type of failure and an Additional Sense Code Qualifier offering further detail. For more information about Additional Sense Code and Qualifier, contact your trained service technician.

Note If a second data drive has a fatal error as described above, the 1275(subsystem$'<will power-down the failed drive and start reporting

.

6500 Disk Array User’s Guide 21020690 B