Archived from groups: microsoft.public.windowsnt.misc (
More info?)
Tritium wrote:
>
> dlw wrote:
>
>> I have about 9 gig worth of files I'm going to move to a new server,
>> if the drive on the new machine is pretty much empty, the copy will in
>> effect defragment the files, right?
>
>
>
> The answer really depends on how the files are copied. If
> they are copied individually, i.e., one file at a time, then
> the degree of defragmentation will be less on the new HD.
> But if the files are copied one folder at a time or one
> logical drive at a time, then the level of defragmentation
> on the new HD increases correspondingly.
>
W2K seems to be much worse on this issue than NT4 was. With
NT4, if you defragged a drive so that all of the free space
was a contiguous block, then if the files were copied one
at a time into that free space, then the first cluster of
free space would be used, then the next, then the next and
so on.
However, with W2K it seem that the first file is given a
*random* location within the free space, then the second
file is given its own random location, and so on. If the
start of each random location begins a block of free space
large enough to hold the file - then that one file will be
OK. However, you end up with a block of free space both
before and after that file - which increases that chance that
the next file will not be started in a large enough block of
free space.
To verify this all you need is a half decent defragging tool
like PerfectDisk or Diskeeper:
1.) Defrag the drive so that all of the free space is consolidated.
2.) Copy a large file onto that drive.
3.) Have the defragger analyze (but do not defrag) the drive. Note
that the file was randomly placed within the free space.
4.) Repeat steps 2 and 3 several times, noting that the free space
becomes increasingly fragmented, and as it does the probability
that each additional file copied onto the drive will be fragmented
increases.
If the idiots at MicroSoft didn't have their heads so far up their
asses, if you copied eighty 1 GB of files one at a time into an 80 GB
block of free space you would end up with no fragmentation for those
eighty new files. However, my simple tests, as outlined above, show
that on average you can only copy 11 files before you start to see
file fragmentation occurring - and free space fragmentation occurs
IMMEDIATELY. At the end of copying the 80 files only 13, on average,
will be unfragmented.
I did that test FORTY times with an 81 GB empty test partition (1 GB to
spare for the MFT), so I have some idea what the heck I am talking about.
However, after the first two times I did not have the defragger analyze
the drive after each file was copied - I only re-analyzed the drive after
all 80 files were copied.
Facebook
Twitter
Instagram