Defects? . . . But Isn't My Drive PERFECT?
No. One of the BIGGEST LIES being told to users of contemporary mass storage systems -- and their operating systems -- is that their drives are "flawless."
THE TRUTH IS that the surfaces of our hard disks are littered with manufacturing variations which adversely affect the surface's ability to retain and represent the data stored there.
Before drives became "intelligent", and able to hide their defective regions, they wore their defects out on their sleeves. The photo below shows a typical RLL drive of that era with a long list of the "known defects" at the time of manufacture.
[click link for image]
Note the "HD-CYL-BYTE" legend showing that the numbers represent
the physical head, cylinder, and byte location of the defects that
were known at the time of the drive's original manufacture.
Why would any manufacturer "advertise" their drive's defects in this manner? Because the purchaser of the drive was required to manually enter this list into the system's low-level formatting program so that the sectors containing those defective spots could be "marked bad" and prevented from ever being used to hold data. As you can imagine, people did not appreciate buying a "new" drive that contained defects. So they'd select the drives with the shortest defect lists, leaving the worst drives for the next guy who didn't know any better.
The instant our drives became intelligent, carrying their own built-in controllers, the very first thing to disappear from the outside of the drive were the defect lists.
An "intelligent" drive's on board brain works to make the drive appear "defect free" by managing a pool of spare sectors and "redirecting" any accesses to the drive's internally defective sectors to the corresponding "reallocated" spare sector. Thus, users all believed that drives became perfect. Also, there was no longer any way to select the "best" drives from the herd. Nor was there any way to tell one manufacturer from another or to know what was really going on inside the drive.
So What's Wrong With Automatic, Internal, Defect Management?
Normally nothing. When everything is working right, automatic internal defect management works like a charm. But the problems arise when things start deteriorating inside the drive, or when newly defective sectors start turning up.
Huh? "Newly" Defective Sectors?
Oh yes. You see, the defects that a drive has when it's first made are just the tip of the iceberg! The mass storage industry has a term for the defects which arise from daily drive use. They are called Grown Defects because defects literally "grow" on the surface of the drive!
Since it is the drive's sworn job to keep a lid on the true defective nature of its storage surfaces, the user and the operating system are the last to know when large areas of the drive begin dying.
The BIGGEST PROBLEM with this autonomous "grown" defect management is that the ONLY TIME the drive can detect that a sector has a problem is when it has trouble reading back YOUR DATA! And, as is discussed on the "Data Recovery" page, the drive gives up on trying to get your data back much too quickly, and then replaces the newly defective sector (which used to be just fine until a defect GREW inside it) with a spare EMPTY replacement!
