How do these bad sectors happen? What are they anyway?
They may happen for the following reasons:
- Read/Write head impacting the patter (rare these days)
- Power failure during a write sequence
- Power surge during a read/write sequence
- Poor quality materials in production of the platters.
The first reason has all but disappeared because of the automatic 'parking' of the read/write heads. (How ever if you drop a hard drive the impact of the drop could jar the read/write heads loose thus causing an impact). It would take a very hard impact to make the read/write head to hit a platter while it is spinning.
The second reason is the main reason you may have bad sectors on the platter. When the power fails if a write sequence is in progress it may skew the sector just enough to cause the sector to be unreadable on the next pass. (A low level format will rebuild all the tracks and sectors bringing a drive back to full capacity).
The third reason is also rare, normally a power surge will damage a hard drive beyond recovery even by the "Data Recover Specialists".
The last reason is why you buy high quality products. Although the process of creating a hard drive platter has been improved and refined over the decades since the first "Winchester" hard drive was introduced there are still companies that use low quality materials and cut corners to increase their profit.
Hard drives made today have a lower failure rate compared to thirty years ago
(with the exception of this one:
WD10EARS). The first hard drives had a 100K hours
MTBF (Mean Time Between Failure) this is an average of all drives produced and how many failed after a given length of use in hours. Western Digital (the inventor of the first small hard drive that would fit in a PC) advertised that their hard drive would last over five years or 50K hours. In fact they lasted over 10 years and over 100K hours. The manufacture comes to these facts by testing, then using statistics of how many drives were produced and how many were returned for repair during the production run to come to the number of hours before a product will fail.
In the early days of the hard drive you could laugh at these claims but as the manufactures came up with better manufacturing techniques and materials the MTBF became more realistic.
Troubleshoot, repair, maintain, upgrade & secure...
Today a hard drive will last well over 300K hours, that is over thirty years. (I have four drives that run 24/7 and are over 10 years old).
Note: The main reason for failure of a hard drive is the electronics and not the metal parts.
A low level format of a hard drive will recover the "bad sector" that was locked out by a higher level drive management program.
When you do a low level format you are effectively erasing the drive (also
the bad sectors), recreating the tracks, then recreating the sectors. It is a fairly long process depending on the program you use.
I used to use a Compaq program that would low level drives (because there
were always bad sectors back in the day) with different formats and interfaces, MFM, RLL, SCSI, and IDE.
The program would do a four pass erase, first all 1's, then all 0's, then 1's again, then all 0's to make the surface clean. Once the erase was complete the next step would be to calculate the number of tracks (a track is a one complete circle of the platter the width of the read/write head, think of a clam pool of water, throw a stone in the middle and watch the waves when the stone goes in to the water, each wave is a complete circle and it is not a concentric circle such as a track on the old vinyl records. Each track is separate from the one either side of it).
When the track step is complete the program creates the sectors.
The standard sector is 512 bytes, it has a start section, a section that tells the drive controller if the sector is part of a cluster and the sector that pointed to this sector, a section for data, then if there are more sectors in the cluster a pointer to the next sector, then the end of sector marker.
The next step is to build the drive table, this lists all the tracks and the sectors in each track. It will also list any sectors that the program could not format (a really bad sector).
The last step before creating a log to list the complete process is to run a special check disk of all the tracks and sectors
looking for bad sectors, this is a more exhaustive check disk than that of say DOS or Windows check disk program.
The last step is to write the log for you to see how many were recovered and how many were not, along with the complete capacity of the hard drive.
You can find low level format programs on the internet, some are free, some cost, along with quality programs to run a through check disk if you want to insuring your data is safe.
Because a SSD (Solid State Drive) does
not have any moving parts, that is platters and read/write heads if you do come
up with 'bad sectors'
when you do a check disk you need to consider backing up your data as soon as
A bad sector in a SSD means the
memory is failing, and because it is then an electronic a
chip is failing and when it fails completely the SSD will be dead, the
same as with a memory module on the motherboard.