Disk Defragmentation and increase of disk space

  • Thread starter Thread starter Roy
  • Start date Start date
R

Roy

Just recently I tried to defrag one of my laptops, and noticed while
the defrag was going on, the disk space increases gradually. This was
contrary to my experience ...As I find it rather strange I stopped the
process.
I would like to ask...
Is this considered normal or just an exception... maybe because as the
files are rearrange the can cover slightly more space...?
TIA
Roy
 
Just recently I tried to defrag one of my laptops, and noticed while
the defrag was going on, the disk space increases gradually. This was
contrary to my experience ...As I find it rather strange I stopped the
process.
I would like to ask...
Is this considered normal or just an exception... maybe because as the
files are rearrange the can cover slightly more space...?
Click to expand...

I wouldn't expect it to be noticable unless you had a ton of large
directories? The files themselves won't change size simply by defragging.
Only the directories could get smaller (if the defrag is done right).

Patrick
========= For LAN/WAN Protocol Analysis, check out PacketView Pro! =========
Patrick Klos Email: (e-mail address removed)
Klos Technologies, Inc. Web: http://www.klos.com/
============================================================================
 
In message
<04c8bd37-c4e8-49bc-91c3-04a26990dc8f@o40g2000prn.googlegroups.com> Roy
Just recently I tried to defrag one of my laptops, and noticed while
the defrag was going on, the disk space increases gradually. This was
contrary to my experience ...As I find it rather strange I stopped the
process.
I would like to ask...
Is this considered normal or just an exception... maybe because as the
files are rearrange the can cover slightly more space...?
Click to expand...

Are you using Vista (or shadow copies on 2003/2008 servers)?

If so, if your defrag app isn't VSS aware it will cause the loss of some
or all system restore points, which could free up significant disk
space.
 
In message
<04c8bd37-c4e8-49bc-91c3-04a26990d...@o40g2000prn.googlegroups.com> Roy


Are you using Vista (or shadow copies on 2003/2008 servers)?

If so, if your defrag app isn't VSS aware it will cause the loss of some
or all system restore points, which could free up significant disk
space.
Click to expand...

I am using win xp sp2,
If its true that it can free significant disk space how come the file
size gradually increase as the defrag went on?...One peculiarity it
was defragging very slowly
 
Roy said:
I am using win xp sp2, If its true that it can free significant
disk space how come the file size gradually increase as the
defrag went on?...One peculiarity it was defragging very slowly
Click to expand...

In the process it has to make copies of many things, if for nothing
else to free up space to hold the final version of some file. When
the process completes, you should have more space. But, if your
disk is nearly full before starting the defrag, it may well fail or
take forever.

Remember that disk defrag is normally a very sensitive process, in
that simple errors in memory buffers can destroy your whole
system. ECC memory is a protection against this. Too many systems
are sold without ECC.
 
Ian said:
.... snip ...


In defragmenters such as Diskeeper, data is verified at the new
location before the old location is released. This gives data
protection, even against power interruption during the defrag
operation, although nothing is absolute.
Click to expand...

Consider how it is verified (which I don't know). If it simply
reads the rewritten file and compares it against the memory buffer,
that won't protect. It will eat up processing time. If it
re-reads the original, and compares that against the rewritten,
that will usually work, but will require a lot more disk i/o.

Actual disk writes from a buffer are already well protected, by use
of a CRC (or other) checksum. The problem is: read a buffer,
buffer drops a bit, write that buffer. This sort of error can
affect almost anything, without the ECC to protect the memory.
 
... snip ...



Consider how it is verified (which I don't know).  If it simply
reads the rewritten file and compares it against the memory buffer,
that won't protect.  It will eat up processing time.  If it
re-reads the original, and compares that against the rewritten,
that will usually work, but will require a lot more disk i/o.

Actual disk writes from a buffer are already well protected, by use
of a CRC (or other) checksum.  The problem is:  read a buffer,
buffer drops a bit, write that buffer.  This sort of error can
affect almost anything, without the ECC to protect the memory.
Click to expand...

How about the default Defragmenter do the behave the same?
 
In defragmenters such as Diskeeper, data is verified at the new
location before the old location is released. This gives data
protection, even against power interruption during the defrag
operation, although nothing is absolute.
Click to expand...

Yup. I accidently switched off the power to my desktop while (while
trying to install satellite speakers on the wall lol) during a
Diskeeper 2008 defrag about a month ago. I was scared I'd lose the
data, but the system booted up, asked to run chkdsk which I allowed,
and then loaded the XP desktop without a problem. Everything was as
before the power outage. Diskeeper is a very well written program!
 
In message
<349d65de-595c-4d75-9c9b-a5c5f244faf4@q26g2000prq.googlegroups.com>
Yup. I accidently switched off the power to my desktop while (while
trying to install satellite speakers on the wall lol) during a
Diskeeper 2008 defrag about a month ago. I was scared I'd lose the
data, but the system booted up, asked to run chkdsk which I allowed,
and then loaded the XP desktop without a problem. Everything was as
before the power outage. Diskeeper is a very well written program!
Click to expand...

Not at all, Diskeeper couldn't screw that one up unless they put a *lot*
of effort into it.

Windows provides a robust defragmentation API which handles journaling
the individual moves to ensure data integrity.
 
In message



Not at all, Diskeeper couldn't screw that one up unless they put a *lot*
of effort into it.

Windows provides a robust defragmentation API which handles journaling
the individual moves to ensure data integrity.
Click to expand...

A quick run through Google reveals that the Windows NT defrag API
dating back to the late 90s (??) was actually a joint effort between
Diskeeper (Executive Software then) and Microsoft. I knew that the XP
defragger was a lame version of some old Diskeeper version, but not
about their role in the Windows defrag API. Interesting indeed.
 
Back
Top