IBM Midrange System DS4000 Series Hardware Manual page 435

This command adds some flexibility to the redundancy check. It is never easy to predict how
long the utility will take to complete and what impact it will have. Therefore, it is advisable to
perform a quick test first by timing how long it takes to complete the check on the first 10 GB
of data. The time to complete execution on the entire logical drive can then be estimated with
some accuracy. If there are performance concerns, then the priority level can be reduced. The
test command looks like this:
check logicalDrive ["LUN01"] parity parityErrorFile="c:\LUN01.parity.txt"
mediaErrorFile="c:\LUN01.media.txt" priority=low verbose=TRUE startingLBA=0
endingLBA=20000000;
Repairing RAID redundancy errors
If RAID parity errors are detected during execution of the logical drive parity check script, then
there is another script command that can be run to fix them:
repair logicaldrive [logicalDriveName] parity parityErrorFile="filename"
[verbose=(TRUE | FALSE)];
So, our usage example looks like this:
repair logicalDrive ["LUN01"] parity parityErrorFile="c:\temp\LUN01.parity.txt"
verbose=TRUE;
This generates the following output in the bottom pane of the script editor while the sectors
with invalid parity are being corrected:
Executing script...
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,032
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,033
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,034
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,035
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,036
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,037
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,038
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,039
Repair Logical Drive Parity called on logical drive: "LUN01" at LBA: 2,040
Script execution complete.
This command corrects the RAID parity information on the logical drive where errors were
detected, but it has no way to determine which disk(s) have valid data and which are
incorrect. Therefore, it is always safer to identify the root cause of the RAID parity corruption
and repair it by forcing a reconstruction of data on the disk most likely to have contributed to
the problem. The script command should only be used when the affected data area is known
to occupy non-critical data or unused space.
Disk media errors
Over time, there will be a degree of deterioration on mechanical and Solid State Drive
devices. The logic on the disks themselves will automatically re-map deteriorating sectors
without the DS4000 or DS5000 controller even being aware. This extends the life and
reliability of the media quite considerably. When enabled, the background Media Scan
function forces a read of every sector in a logical drive over a preset period.
417
Chapter 7. Advanced maintenance, troubleshooting, and diagnostics