The author here explains the meaning of the parameters in the "health" test (for reference only, because the meaning of the original data is confidential by the hard disk manufacturers, and different manufacturers may be different).
Let's take a look at the attribute table of S.M.A.R.T:
South M.A.R.T contains many attributes, and each attribute value has only two meanings, exceeding the threshold and not exceeding the threshold. Once an attribute value exceeds the threshold, it means "the hard disk is dying". The thresholds of these attributes are indicated in the S.M.A.R.T information of each hard disk, and the thresholds of different manufacturers are different. There are two contrast relationships between different attribute values and thresholds, which should be greater than the threshold and less than the threshold.
Here are some more important attributes that I think need to be explained:
(0 1) read error rate (should be less than the threshold)
When there is an error reading data from the disk surface, record the hardware reading error rate. Any data greater than 0 indicates that there is a problem with the disk surface or reading or writing column headers. If the minimum difference of this item is very low, it means that there are probably bad tracks on the hard disk. If this value is not normal, it may be that some second-hand hard disks have been repaired or blocked.
(05) The reallocated sector count for remapping (should be less than the threshold).
Count value of remapped sectors. When the hard disk finds errors in reading, writing or checking, it will remap this sector and transfer the data to a specially reserved free area, which is called remapping sector. In other words, the hard disk can't find "bad blocks" through detection, and all the bad blocks are hidden in the remapped sectors, which will of course slow down the reading and writing speed.
(09) Power-on time count, POH (the smaller the better, usually there is no threshold)
Hour count in power-on state. This value represents the total hours when the hard disk is powered on, but the unit of this value is different from different manufacturers, and it is also minutes and seconds. The newly bought hard disk should be less than 10 hour, otherwise it is possible. . . . Hey hey.
(0A) Rotation retry count of motor (should be less than threshold)
The number of retries that the motor tried to start. This property stores the total number of motor start attempts to reach the standard speed, that is, the first start failed to reach the standard speed. There may be a problem with the high-value circuit or reserved area of the hard disk, indicating that there is a problem with the mechanical system of the hard disk.
(0C) Power cycle count (the smaller the better, there is usually no threshold)
This attribute represents the number of on/off cycles of the whole hard disk, that is, the number of switches.
(C2) Temperature (the smaller the better, there is usually no threshold)
Current internal temperature. Generally, it should not exceed 45 degrees Celsius.
(C4) Re-map the event count and redistribute the event count (which should be less than the threshold).
Count value of remapping operation. The attribute value indicates the total number of attempts to transfer the data of the remapped sector to the free area. Both successful and unsuccessful transfers will be counted.
(C5) The current pending sector count of the sector to be mapped (should be less than the threshold).
The number of "unstable" sectors, that is, the number of sectors waiting to be mapped. If the unstable sector is subsequently read and written successfully, the value will be reduced and the sector will not be remapped. A sector reading error will not lead to remapping, and only when the writing fails will the sector be remapped. This value is sometimes problematic because cached writes do not remap sectors, and only direct reads and writes actually write to disk. If this value is not normal, it may indicate that the hard disk is about to hang up. . . .
(C6) The sector count that cannot be corrected (which should be less than the threshold) cannot be corrected.
The total number of uncorrectable errors when reading and writing sectors. An increase in this value indicates a defect in the hard disk or a mechanical system problem.
(C7) UltraDMA CRC error count of direct memory access check error count (should be less than the threshold).
Errors in data transmission of interface cable discovered through interface cyclical redundancy check (ICRC).
(C8) Write error rate (should be less than the threshold)
The total number of errors that occurred while writing sectors.
If one or more of the above items "turn red", the problem may be more serious, so deal with it quickly. . . Otherwise, just ... . Not far. . .
Reference address/14/86 _136852.html