User manual

Using S.M.A.R.T. Disk Monitor 165
SANTOOLS® is registered in US Patent and Trademark Office No 3,107,854 All rights reserved.
clueless as to what you are asking for. Request an engineer that understands programming).
If you have neither the time, desire, or resources to chase whether or not it is possible to report something, contact
SANtools directly. ([e-mail]). For additional fees, we will be glad to play detective and provide you with
a script to report what you desire. We have non-disclosure agreements with most peripheral manufacturers, so we
can typically get the programming information required to meet your needs.
If you cannot wait, you must make a manual entry into the threshold file that describes the log page, parameter (or
offset), description, and byte length. We will assume you have obtained this information from the technical support
department of the device's manufacturer. Make an entry in the configuration file as documented above. Be sure to
note that the log information must be entered in hex, as that is how the manufacturer documents these settings.
7. Important note - feature change in 1.23B
With the point-release 1.23B, we made an important change in how this feature works. Before this release, the
threshold monitoring was combined with S.M.A.R.T. monitoring. The program would scan all devices, enable SMART
polling at whatever interval you defined, and concurrently do threshold monitoring at the desired intervals.
We improved the behavior by removing this additional logic. Now if you want to monitor thresholds, that is all that will
be monitored. No I/Os other than log sense I/Os (and a standard Test Unit Ready command) are sent to the device to
obtain the data you desire.
8. Important note - syntax change in 1.25 (Windows only)
With the changes required to the device naming convention that were necessitated by Microsoft's new SCSI drivers,
we were forced to change the syntax of the file for the Windows distribution. The syntax now matches the UNIX
format. If you upgrade to version 1.25 and attempt to run a configuration file that was built with a prior release of the
program, the program will detect that you are using an older-formatted file and reject the command. The quickest and
easiest way to convert the file is to edit the file with your favorite text editor and replace the first three parameters
(which originally contained the SCSI Channel, ID, and LUN) with the device path name as shown in the sample above
.
1.48 Verify Data
This feature was added in release 1.41. It instructs the selected disk(s) to invoke the built-in SCSI verify function.
This function is built into most disk drives and runs very quickly with near-zero host overhead. Feel free to use this on
as many drives as you wish concurrently. (You must, however, run multiple instances of the software as the program
will lock up until the current drive completes the process.
The -verify command is supported on SCSI, SAS, Fibre Channel disks under all operating systems. It is also
supported on ATA/SATA disk drives under Windows. (If you have ATA or SATA disks on other operating systems,
then check with us to see if this command is ported to your operating system.
Benefits of running SCSI Verify Function
· The -verify function runs inside of the drive firmware, so there is near-zero host overhead.
· The -verify is the fastest technique possible for the disk drive to make sure that there are no bad disk blocks.
You can verify as many disks as your hosts can support concurrently.
· Blocks go bad 24x7. This command will tell you if you have any bad blocks, regardless of whether or not the block
is being used by a file, before the operating system asks for the data. Once you know where you have corruption,
you can react accordingly. Remember, if you have RAID5 and lose a disk drive, but have a bad block on a
surviving disk, then you have 100% data loss for that chunk. Furthermore, some RAID controllers will fail a rebuild
in this situation and you could very well be left with a RAID system that will not repair itself.
Syntax
-verify -scrubv (the -scrubq makes output verbose so bad blocks and percent complete will be reported as the
drive progresses).
Example - Verifying a SCSI Drive
161