What is it with defrag'ing?

[Josef Stalin]

Comrades,

It seems that in every 10'th post, someone comments about the need to defrag
a disk to solve some problem. Some recent examples:

"I have a Windows 2000 pro station that is giving me some trouble. I have
defragged the drive ..."

"when I run a Mcafee virus scan it gets half way thru and the computer goes
and reboots. This happens when I run spybot software too. I ran the
defragmenter and it runs ok, ..."

"Would need a lot more info - anything in your event logs? Have you
defragged recently?"

"Although, running a scan disk, it hangs up on 4 bars and will not scan my
machine. I can defrag tho."

I don't get it. It's like getting washing your car when the engine is
running poorly. The people who sell the defraggers are doing one great job
in overstating the merits of the products.

I've said it before, and I'll say it again: a defrag'er will, in general,
give you only incremental improvement in performance. More often than not,
simply a perceived improvement. If you're experiencing difficulties with
your system a defrag'er will never, ever solve your problem.

Uncle Joe

 

[Pegasus \(MVP\)]

Comrades,

It seems that in every 10'th post, someone comments about the need to defrag
a disk to solve some problem. Some recent examples:

"I have a Windows 2000 pro station that is giving me some trouble. I have
defragged the drive ..."

"when I run a Mcafee virus scan it gets half way thru and the computer goes
and reboots. This happens when I run spybot software too. I ran the
defragmenter and it runs ok, ..."

"Would need a lot more info - anything in your event logs? Have you
defragged recently?"

"Although, running a scan disk, it hangs up on 4 bars and will not scan my
machine. I can defrag tho."

I don't get it. It's like getting washing your car when the engine is
running poorly. The people who sell the defraggers are doing one great job
in overstating the merits of the products.

I've said it before, and I'll say it again: a defrag'er will, in general,
give you only incremental improvement in performance. More often than not,
simply a perceived improvement. If you're experiencing difficulties with
your system a defrag'er will never, ever solve your problem.

Uncle Joe

Well said. In most cases, defragging a hard disk is like peeing
in your wetsuit: It gives you a warm feeling but nobody notices.
There is a real problem, however: Defragging involves moving
countless clusters of data from here to there. Doing so involves
a small but finite risk of something going horribly wrong. In these
newsgroups we sometimes see the result of this mostly
unnecessary operation.

 

[GO]

Well said. In most cases, defragging a hard disk is like peeing
in your wetsuit: It gives you a warm feeling but nobody notices.
There is a real problem, however: Defragging involves moving
countless clusters of data from here to there. Doing so involves
a small but finite risk of something going horribly wrong. In these
newsgroups we sometimes see the result of this mostly
unnecessary operation.

I've often wondered what the merits were of running defrag tools nowadays
(especially on ntfs partitions). I have no doubts that back when I was
running Win9x that defragging was definately beneficial but I have yet to
see a need to do it now (although I still do it on occasion . I've
personally witnessed a WinNT 4 box that hadn't been defragged in years (and
it was a horrible mess in the "eyes" of the defragger) and after several
runs with one there was no noticable improvement.

 

[Guest]

Not that I'm a techie of any kind so you guys will tell me if I'm barking up the wrong tree, but 'visible improvement' aside, I always thought one point of defragging was to gather files together so as to minimise the HD movements back and forth 400 times over the disk in order to read a single file, and thus to give it an easier task and thus hopefully minimise future issues with harddrives breaking or damaging sectors etc.

????

/peter

 

[Pegasus \(MVP\)]

Not that I'm a techie of any kind so you guys will tell me if I'm barking

up the wrong tree, but 'visible improvement' aside, I always thought one
point of defragging was to gather files together so as to minimise the HD
movements back and forth 400 times over the disk in order to read a single
file, and thus to give it an easier task and thus hopefully minimise future
issues with harddrives breaking or damaging sectors etc.

????

/peter

What you say is correct but I have yet to see an credible relationship
between disk longevity and frequent defragging in a workstation environment.
The vast majority of disks are replaced because they become too small.

 

[Guest]

To Pegasus (MVP):
I can tell you are good (you'll find what I'm about to say). I was browsing a bit, waiting for my post to appear, saw you little tit-a-taa with that one guy. Cute.

I posted about an oddity titled "Check Disk for Errors Log - Oddity?", and what contents of "bootex.log" (output of Chkdsk) means.

I do defrag on occasion (never to "fix" anything), but maybe to get that little extra perception boost.

I wonder if this output is a result of frequent defrags. (Defrag gives message you have 22% free space, but only 12% is available for Defrag to use. Defrag must have 15% of useable space to..."

I asked about this once, everyone here told me don't worry about that, start the defrag, then go to bed.

Or is the inconsistancies (see my post) anything with Compressed files (I compress some folders not entire drive). Only 1 partition on one drive.

Thanks
Michael

 

[Dan Seur]

An interesting thread so far. I believe with a well-configured system
(fast HDDs, not full, pagefile well-placed, savvy partitioning) defrags
generally yield wristwatch-imperceptible perf improvement. But as with
icediving after coffee, where peeing in the suit can be a very good
idea, and with Congressional budgeting, where as some immortal observed
"a billion here, a billion there...pretty soon you're talking real
money", I believe there are circumstances where good housekeeping does
involve defragging (among other regular chores). I believe this applies
to NTFS-structured partitions as well, although NTFS is vastly improved
over some prior MS formats in this regard.

Defrags are never a cure for software or hardware problems, but I
believe there's a kind of "knee" in the performance curve that defrags
can help one avoid. And many users don't have particularly well
configured systems, as is made plain in these NGs. And posters here are
savvy enough to find their way here; imagine what goes on elsewhere.

As ever, I'm open to the notion that I'm tilting at windmills.

 

[Bob]

What you say is correct but I have yet to see an credible relationship
between disk longevity and frequent defragging in a workstation environment.
The vast majority of disks are replaced because they become too small.

I've seen measurable improvements in application loading for large
programs after defragging. However, if you are running multiple
applications concurrently, the advantages are probably minimal
since the disk has a lot of traveling to do anyway.
Disk Performance will always be better when it can read with
minimum movement but the key question is whether single large file /
large application read disk performance is an issue for you in terms
of performance.

 

[Greg Hayes/Raxco Software]

Michael,

"I wonder if this output is a result of frequent defrags. (Defrag gives
message you have 22% free space, but only 12% is available for Defrag to
use. Defrag must have 15% of useable space to..."

Frequent defrags do not result in file system issues

Regarding your 22% free space but only 12% is available.

Free space on NTFS drives is located in 2 different areas - INSIDE of the
MFT Reserved Zone and OUTSIDE of the MFT Reserved Zone. Microsoft's defrag
APIs under NT4 and Windows 2000 do not let defragmenters use the free space
inside of the MFT Reserved Zone. That is why you are getting this
message...

- Greg/Raxco Software
Microsoft MVP - Windows File System

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a
commercial defrag utility, as a systems engineer in the support department.

Want to email me? Delete ntloader.

To Pegasus (MVP):
I can tell you are good (you'll find what I'm about to say). I was

browsing a bit, waiting for my post to appear, saw you little tit-a-taa with
that one guy. Cute.

I posted about an oddity titled "Check Disk for Errors Log - Oddity?", and

what contents of "bootex.log" (output of Chkdsk) means.

I do defrag on occasion (never to "fix" anything), but maybe to get that little extra perception boost.

I wonder if this output is a result of frequent defrags. (Defrag gives

message you have 22% free space, but only 12% is available for Defrag to
use. Defrag must have 15% of useable space to..."

I asked about this once, everyone here told me don't worry about that,

start the defrag, then go to bed.

Or is the inconsistancies (see my post) anything with Compressed files (I

compress some folders not entire drive). Only 1 partition on one drive.

 

[Guest]

All

On a slightly different note, is it any worth the while defragging with Norton's Speed disk rather than the Windows defragmenter? When I was using win98 I always used Norton since one could specify files to place first and last and/or frequently accessed(/modified files first or last etc etc. I kind of liked that feature but never knew if it actually did anything good. Now with win2000 SP4 Speed disk says it has not been tested with SP4 so I don't want to use it anymore, and I am back to the Windows defragmenter. Will it do the job good enough for me (home user)? Or do I need to look for when Norton has updated to support SP4 and switch when they do or can I stick to Windows defrag

best

P

 

[Josef Stalin]

All,

On a slightly different note, is it any worth the while defragging with

Norton's Speed disk rather than the Windows defragmenter? When I was using
win98 I always used Norton since one could specify files to place first and
last and/or frequently accessed(/modified files first or last etc etc. I
kind of liked that feature but never knew if it actually did anything good.
Now with win2000 SP4 Speed disk says it has not been tested with SP4 so I
don't want to use it anymore, and I am back to the Windows defragmenter.
Will it do the job good enough for me (home user)? Or do I need to look for
when Norton has updated to support SP4 and switch when they do or can I
stick to Windows defrag?

best,

I think the point of my original post has been lost.

I'm not against defragging and I believe everyone should use it regularly.
In fact, I use it several times a week. Probably because I might have OCD
than it does any real good. I'm sure that any good defragger is
sufficient.

My point of contention is that all too often people believe it's a tool to
fix disk or system problems.

 

[Major Malfunction]

Norton's Speed disk rather than the Windows defragmenter? When I was using
win98 I always used Norton since one could specify files to place first and
last and/or frequently accessed(/modified files first or last etc etc. I
kind of liked that feature but never knew if it actually did anything good.
Now with win2000 SP4 Speed disk says it has not been tested with SP4 so I
don't want to use it anymore, and I am back to the Windows defragmenter.
Will it do the job good enough for me (home user)? Or do I need to look for
when Norton has updated to support SP4 and switch when they do or can I
stick to Windows defrag?

I think the point of my original post has been lost.

I'm not against defragging and I believe everyone should use it regularly.
In fact, I use it several times a week. Probably because I might have OCD
than it does any real good. I'm sure that any good defragger is
sufficient.

My point of contention is that all too often people believe it's a tool to
fix disk or system problems.

Amen to your reply! Let me further clarify that I'm on your side: defrag is
not a panacea, but it *can* help performance under certain conditions. I
usually use defrag after a major archive operation. I find removing garbage,
and subsequently increasing free disk space, effectively increases
performance and following this with a defrag further reduces the performance
lag.

Case in point: I have a Toshiba laptop with a Celeron 400 and 192 Mb of RAM.
hardly a race horse, but it is a traveling briefcase (read file server for
one user and occasional remote internet connection for email), so more
horsepower is not really needed. Disk space was reduced to 4% free space and
the system crawled to the point of taking 15 seconds to execute what was
supposed to be a 5 second slide change in a PDF slide show. I removed unused
/ unneeded files and brought the free space to a more respectable 20%. Slide
changes were reduced to 8 seconds by freeing the swap space. A defrag
operation was performed and the slide show is now performing as it supposed
to. YMMV, but removing unneeded garbage is always a good suggestion.

 

[William W. Plummer]

tool
Amen to your reply! Let me further clarify that I'm on your side: defrag is
not a panacea, but it *can* help performance under certain conditions. I
usually use defrag after a major archive operation. I find removing garbage,
and subsequently increasing free disk space, effectively increases
performance and following this with a defrag further reduces the performance
lag.

Case in point: I have a Toshiba laptop with a Celeron 400 and 192 Mb of RAM.
hardly a race horse, but it is a traveling briefcase (read file server for
one user and occasional remote internet connection for email), so more
horsepower is not really needed. Disk space was reduced to 4% free space and
the system crawled to the point of taking 15 seconds to execute what was
supposed to be a 5 second slide change in a PDF slide show. I removed unused
/ unneeded files and brought the free space to a more respectable 20%. Slide
changes were reduced to 8 seconds by freeing the swap space. A defrag
operation was performed and the slide show is now performing as it supposed
to. YMMV, but removing unneeded garbage is always a good suggestion.

defrag does not make a noticeable change in performance. This is because
Windows uses a lot of RAM for a disk cache. (OK, that doesn't help things
the first time you access each file.) Now, you said your laptop is severely
lacking in RAM -- that explains the performance difference.

 

[Greg Hayes/Raxco Software]

In only 1 instance have I seen an instance where using Microsoft's defrag
APIs resulted in file system corruption. That was in Windows 2000 RC2 and
it was a very specific instance - FAT32 drive greater than 4GB. Other than
that, there have been no instances reported or "fixed" by Microsoft where
use of their defrag APIs resulted in file system corruption.

- Greg/Raxco Software
Microsoft MVP - Windows File System

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a
commercial defrag utility, as a systems engineer in the support department.

Want to email me? Delete ntloader.

 

[Greg Hayes/Raxco Software]

"credible relationship between disk longevity and frequent defragging in a
workstation environment. The vast majority of disks are replaced because
they become too small."

I would agree with this. Disk longevity is primarily due to how good the
disk hardware is. Defragmenting can't improve hardware

- Greg/Raxco Software
Microsoft MVP - Windows File System

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a
commercial defrag utility, as a systems engineer in the support department.

Want to email me? Delete ntloader.

 

[Greg Hayes/Raxco Software]

"defrag does not make a noticeable change in performance. "

Somebody forgot to notify Microsoft of this. Fragmentation is of such
importance that Windows XP attempts to mitigate the effects of fragmentation
automatically. Most people aren't even aware of the steps that Windows XP
takes in this regard.

- Greg/Raxco Software
Microsoft MVP - Windows File System

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a
commercial defrag utility, as a systems engineer in the support department.

Want to email me? Delete ntloader.