SCF Reference Manual for the Storage Subsystem (G06.28+, H06.05+, J06.03+)

Considerations for Changing the Speed of a Revive Operation
• The speed of all future revive operations is also changed.
• During system installation, the system administrator should tailor the REVIVEPRIORITY and
REVIVERATE attributes for your system. The default values provided by SCF are acceptable
for most environments:
◦ REVIVEPRIORITY 50
◦ REVIVERATE 100 second between copies
The default values minimize potential interference with system performance but could
result in revives that take too long to finish. (The longer the revive operation takes, the
longer your mirrored disks have dissimilar data.)
• To speed up the revive operation (even though this change might slow system performance),
increase the REVIVEPRIORITY value and/or increase the REVIVERATE value.
• If you change these values while a revive operation is in progress, the disk process does not
restart the revive operation from the beginning but continues from the point at which you
entered the new values.
For more information, see REVIVERATE.
Example of Changing the Speed of a Revive Operation
This command establishes a revive priority of 60 and specifies that 90 megabytes of data be
revived between preemption checks while a revive operation is in progress:
-> ALTER $DATA01, REVIVEPRIORITY 60, REVIVERATE 90
Stopping a Revive Operation
It is seldom necessary to stop a revive operation (with a STOP DISK command on the disk being
revived) unless you want to force the revive operation to restart from the beginning.
You might want to adjust a revive operation if:
• System performance is degraded. See “Changing the Speed of a Revive Operation” (page 99).
• A media error has occurred, causing a defective sector.
If you have enabled automatic sector reallocation, the system spares the sector and the revive
operation resumes. If automatic sector reallocation is disabled, see “Temporarily Stopping a
Revive Operation” (page 101).
Nonfatal checksum errors have stalled a revive operation or put it into a loop. Symptoms are:
◦ The revive interval in a STATUS, DETAIL display has become large.
◦ The Current Logical Sector remains unchanged.
◦ EMS messages are generated. Research these messages with the Event Management
Service (EMS) Analyzer User’s Guide and Reference Manual.
The revive operation continues to retry at the current address until it is successful,
suspended, or stopped.
After each retry, the revive interval doubles in length. In this way, the retries do not
consume system resources or produce too many operator messages. The interval continues
to double up to a maximum of one hour; thereafter retries occur indefinitely. If a retry
succeeds, the revive interval returns to its original configured value.
See “Temporarily Stopping a Revive Operation” (page 101).
100 Managing Disks