Disk Defrag

[rick s]

Someone told me that the more you run disk defrg the better chance of
destroying your hard drive. Is this true?

 

[VanguardLH]

in

Someone told me that the more you run disk defrg the better chance of
destroying your hard drive. Is this true?

The more activity, the more wear, just like the more you use your
bicycle the more likely the chain will break. So, do you not use your
hard disk at all between these defrags? You power up, do a defrag, and
then power down until the next time you defrag?

 

[Pegasus \(MVP\)]

Someone told me that the more you run disk defrg the better chance of
destroying your hard drive. Is this true?

There is a small risk of damaging your file system while
defragging, e.g. when you suffer a power failure right in
the middle. I think defragging once a month is plenty. Any
more often would make no measurable difference but
would increase the risk.

 

[Unknown]

Absolutely untrue.

 

[Colin Barnhorst]

It is not so much that you're running defrag as it is disk usage. Anything
that heavily accesses the disk can eventually cause wear and tear on the
mechanisms sufficiently to shorten disk drive life (even if slightly) when
run excessively. You can get the same effect by constantly running chkdks
or other disk utilities far more often than needed. But don't let that
slight concern prevent you from running defrag appropriately (when the
defrag analyser says you should defrag). These are reasons why folks should
maintain a good backup plan. Hard drives do eventually just wear out even
when using best practices. I have some that have lasted many years and I
have had some that only made it a year.

 

[Bill in Co.]

in


The more activity, the more wear, just like the more you use your
bicycle the more likely the chain will break. So, do you not use your
hard disk at all between these defrags? You power up, do a defrag, and
then power down until the next time you defrag?

This begs the question: just how much is the MTBF reduced by running a
defrag more often? I'm not sure anyone has collected the data to really
know. I bet its impact is minimal, however, in practical terms.

 

[VanguardLH]

This begs the question: just how much is the MTBF reduced by running a
defrag more often? I'm not sure anyone has collected the data to really
know. I bet its impact is minimal, however, in practical terms.

Like begging the question as to how much shorter is the MTBF if there is
less physical RAM so more pagefile space must be used.

I know of users that like to use the flash memory thumb drives for
pagefile space not realizing how often the pagefile gets used. Flash
memory does wear out due to oxide stress. There is logic available on
the drive to mask out bad areas but eventually the reserve gets used up
and the thumb drive suddenly and catastrophically dies taking the
pagefile with it and any changes saved there during the current Windows
session.

 

[VanguardLH]

There is a small risk of damaging your file system while
defragging, e.g. when you suffer a power failure right in
the middle. I think defragging once a month is plenty. Any
more often would make no measurable difference but
would increase the risk.

I believe the use of journaling with NTFS compensates for that loss, and
also if the content of the hard disk's PCB write buffers aren't flushed
to the platters, too. Another reason to use NTFS rather than FAT. NTFS
can use its journal log to determine how to return the disk to a
consistent state upon recovery.

 

[Pegasus \(MVP\)]

I believe the use of journaling with NTFS compensates for that loss, and
also if the content of the hard disk's PCB write buffers aren't flushed
to the platters, too. Another reason to use NTFS rather than FAT. NTFS
can use its journal log to determine how to return the disk to a
consistent state upon recovery.

I am aware of this mechanism. It appears to work most of the time
but once every so often we get a post in this newsgroup that reports
a thrashed filing system after an interrupted defrag.

 

[Bill in Co.]

Like begging the question as to how much shorter is the MTBF if there is
less physical RAM so more pagefile space must be used.

Yup Another interesting question.

I know of users that like to use the flash memory thumb drives for
pagefile space not realizing how often the pagefile gets used.

Sounds like a BAD idea (per below).

Flash memory does wear out due to oxide stress. There is logic available
on
the drive to mask out bad areas but eventually the reserve gets used up
and the thumb drive suddenly and catastrophically dies taking the
pagefile with it and any changes saved there during the current Windows
session.

Right. We just had a discussion on that in the other group, and how the
flash memory lifetime is limited to begin with, and running defrag on it can
significantly reduce its lifetime. (Someone mentioned a figure of only
1000 erase/writes for each cell, but that the overall memory lifetime, from
a user's standpoint, is increased (by the chip logic you mentioned), so that
the end result is the average lifetime figure is more like 10,000 - 100,000
erase/write cycles.

 

[Bill in Co.]

I am aware of this mechanism. It appears to work most of the time
but once every so often we get a post in this newsgroup that reports
a thrashed filing system after an interrupted defrag.

Which indicates the journaling actually failed? Wow. Well, maybe its
not all that unexpected.

Admitedly the overhead of NTFS has it advantages here, over FAT. I'll
have to concede that.

 

[Unknown]

I would venture to say there is much more wear and tear on a fragmented file
if you don't defrag.

 

[Colin Barnhorst]

Yes. The heads have to move all over the place to link together a
fragmented file and the system has to do this all the time whereas a defrag
operation may take an hour and result in much reduced head movement for
weeks.

 

[Gerry]

Rick

I have always used Disk Defragmenter a fair bit. It is over 10 years
since I had my only hard disk failure. The disk was under warranty and
replaced free of charge. New disks are just as likely to fail as old
ones.


--



Hope this helps.

Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~

 

[db.·.. >]

there is one thing you
should keep in mind
and that is if your disk
is pretty filled up with
data and you have very
low disk space, then there
is a lot of burden on
the hard drive.

but if you have lots of
harddisk space, then
run the defrag as much as
your little heart desires.

eventually you will
come to realize that
defragging takes time
and is not always
needed especially with
large capacity disks.

 

[Fiddler]

That as bad as being told not to turn the pc off because it will break it

 

[Fiddler]

That's as bad as being told that turning the pc on and off will wreck
it(wear it out).
IF that is so then why have power buttons on computers?

 

[RJK]

Well said ! ...as I read down through this thread, I was poised to make
exactly this point, and found that you already have !
This would also have been my contention i.e. a regularly defragmented hd
results in significantly reduced hard-disk "voice coil type" actuator
activity across the next few weeks.

regards, Richard

 

[Bill Sharpe]

That as bad as being told not to turn the pc off because it will break it

Defrag only when necessary.
Now define "necessary." <g>

One solution is to use the command line:
defrag <drive> -a -v (-a for analyze only, -v for verbose)
which will quickly tell you how defragmented your drive is and the MS
suggestion as to whether defragging is recommended or not.

Bill

 

[Gerry]

Bill

You can have two different scenarios. A preponderance of greatly
fragmented files that you never access and greatly fragmented files that
you frequently access. In generall you are more likely to see the second
scenario. The numbers game is not a good guide on the need to defragment
as it fails to distinguish between important files and those which do
not matter.


~~~~


Gerry
~~~~
FCA
Stourport, England
Enquire, plan and execute
~~~~~~~~~~~~~~~~~~~