AdminXP.com: SMART (Self-Monitoring and Reporting Technology)

SMART (Self-Monitoring and Reporting Technology)

Self-Monitoring and Reporting Technology - technology designed to monitor the disk's activities and performance, and to warn against possible data loss in case some changes of parameters has been made.
To be able to use this technology in its completeness, there has to exist the support of the motherboard (especially in the BIOS system), the disk(EIDE and SCSI) and the OS on your computer.
MB - Enabling and disabling of the SMART functions you can find and use in the BIOS system (Advanced BIOS Features - HDD S.M.A.R.T. Capability - Enabled/Disabled).
Disk - Most newer disks already have the SMART functions implemented in the disk's firmware. The SMART technology differs in the amount of functions according to particular producers. Before the SMART technology was accepted as standard the particular SMART functions had been presented under different names, e. g. Predictive Failure Analysis (PFA) by IBM or Drive Failure Predication (DFP) by Compaq.
Operating System - Without the OS support the possible problems will not be reported to the user. More recent Microsoft Windows systems contain this support. For advanced monitoring there exist a great number of SW: an interesting program called ActiveSmart you can find on www.ariolic.com/activesmart or EZSmart by Storagesoft, Norton System Doctor's SMART sensor by Symantec.
When an error has been indicated, it is advisable to back up the data properly, use a proper disk diagnostics and if needed to change the disk.

Monitored features

The features are set by the producer. Each feature can have a maximal or a minimal value that may not be exceeded. If this happens, a warning against possible problems will appear. The amount and the type of the features depend on the producer - there are monitored up to several dozens of features in some disks.
What can be monitored? E. g. the disk's performance, bad sectors, calibration, CRC errors, disk spin up time, distance between the head and the disk, temperature, features of medium, heads, motor or servomechanism.
The main task of the monitoring is to monitor different features of the disk and in case of some larger deflection to predict potential problems. These problems are presupposed to appears as a result of wear, material defect or some other faults that can be detected partially in advance. So, when the disk does not work for 100%, it is possible that there is something wrong with it. The SMART technology cannot predict the errors occured in consequence of bad handling, an error caused by the user, HW equipment or some other errors that cannot be predicted. Practically, the percentage of unpredictable errors is very low.
Further information you can find on the webs of particular producers - e. g. IBM