Compaq StorageWorks Data Replication Manager HSG80 ACS Version 8.5P Operations Guide

62 Compaq StorageWorks Data Replication Manager HSG80 ACS Version 8.5P Operations Guide
Failure of One Member in a Dual Redundant Controller Pair on page 66
Failure Modes of a DT System in Normal Operation on page 67
Failure of Both Fiber Optic Cables or Switch on page 66
Failure at Target Site after Failover on page 69
HSG80 Array Controller Operating Characteristics
The HSG80 array controller has certain characteristics that may become evident when
used in a Data Replication Manager solution. The following sections will help you
understand these characteristics and educate you on how to respond to them.
Forced Errors Detected During Copy
A forced error is a data bit indicating that a corresponding logical data block contains
unrecoverable data. If a read request from the initiator to the target encounters a forced
error during a full copy, then the data in that block will be copied to the target and marked
with a forced error. These forced errors are then reported to the host and reappear each
time the block is read. The file containing the forced error qualifier should be restored
from a known good backup.
Refer to the HSG80 Array Controller ACS Version 8.5 Maintenance and Service Guide for
complete details on how to recover from a forced error situation.
Read Errors Detected During Full Copy
During normal operation, an unrecoverable error is reported to the host, the offending
block is re-vectored, and the new block is marked with a forced error. During a full copy,
however, the handling is slightly different because the block that is unrecoverable may not
be within normal file system space. Therefore, the controller will terminate the copy and
report the event.
Unrecoverable read errors on the source member will terminate the copy and send a fault
management report to the host. Refer to the HSG80 Array Controller ACS Version 8.5
Configuration and CLI Reference Guide for more information on how to interpret these
logs.