[ The PC Guide | Systems and Components Reference Guide | Hard Disk Drives | Hard Disk
Performance, Quality and Reliability | Hard Disk Quality and
Reliability | Hard Disk Quality and Reliability Issues ]
Failure Modes and Failure Determination
There are a large number of different ways that a hard disk can fail; these are
normally called failure modes, and determining that a failure has occurred--and
why--is called failure determination. The most "famous" way that a hard
disk can go is via the infamous head crash. In
reality, head crashes are the "airline disasters" of the hard disk world: they
are dramatic, well-known and feared, but actually responsible for a small percentage of
total drive problems.
A comprehensive list of failures that can cause a drive to fail would be as long as
your arm, and there's really no point in compiling one. This short list will give you a
flavor for the types of failures that occur:
- Mechanical Failures: These include component failures related to the
mechanics of the disk. Problems of this type usually relate to the spindle motor or
bearings, such as motor burnout, "stuck" bearings, excessive heat, or excessive
noise and vibration. Actuator problems would also fit into this general category.
Unsurprisingly, mechanical failures of hard disks comprise a large percentage of total
- Head and Head Assembly Failures: The infamous head crash fits in here,
as do other problems related to the heads: improper flying height, head contamination,
defects in head manufacture, excessive errors on reads or writes, bad wiring between the
heads and the logic board. These too comprise a large percentage of total failures.
- Media Failure: This class of problems relates to trouble with the
platters and the magnetic media, formatting, servo operation and the like. This would
include drives that fail due to read or write errors, poor handling, scratches on the
media surface, errors in low-level formatting, etc. They are relatively uncommon.
- Logic Board or Firmware Failures: These are problems related to the
drive's integrated logic board, its chips and other components, and the software routines
(firmware) that runs it. Again, problems in this area of the disk are fairly uncommon
compared to the other categories.
Determining that your drive has a failure isn't always the simplest thing to do. Some
failures manifest themselves rather clearly and obviously; others are quite a bit more
subtle. Here are some of the most common signs that your hard disk may be
failing--not definitely, but possibly. (Remember that sometimes apparent drive failures
turn out to only be problems with configuration, drivers or the like... also see the Troubleshooting Expert section on hard disks):
- Errors: If the drive is returning read errors or "sector not
found" errors, that is an obvious sign of a problem. Sometimes errors appear at a
"higher level"; for example, if you experience repeated corruption of the
Windows registry, this could be due to a drive problem (though a hundred other things
could cause it as well.)
- Changed Behavior: If the drive has suddenly changed its behavior in a
dramatic way due to nothing you (believe) you have done, this may be a bearer of bad
tidings. This is especially true of noise: if the drive always made a particular noise
then that noise is probably normal for the drive type (some models make certain clunks or
clangs when they start or shut down). But if the drive starts making new noises
then you should be concerned. If the drive starts vibrating more than it used to, or spins
up and down a lot more than before, these are also potential warning signs.
- Scary Noises: The above said, there are certain noises a drive should never
make, even when new. If you hear something scraping as the drive turns on, or the drive
makes repeated clunking sounds during operation, or emits a very high-pitched whine, then
there may well be a problem.
- Drive Not Recognized: If the drive is not recognized at all by the BIOS
of a motherboard to which it is connected where it was recognized by that BIOS in the
past, this is a sign of a failed drive. (Non-recognition of a new drive can be due to
BIOS compatibility issues.)
- SMART Alarm: If your hard disk and system support the SMART feature and you receive a SMART alert, consider this
fair warning that the drive may have a problem.
Warning: If you suspect
that your hard disk is failing, take action. Back up your data immediately, and
contact the drive manufacturer for instructions. See the section on
warranty issues for more details.
Next: Software Causes for Reliability Issues
Home - Search
- Topics - Up