to defrag or not???

[spinejam]

do you guys defrag your hd? if so, do you notice any benefits or is it just psychological?

 

[NXIL]

http://www.maximumpc.com/artic...disk_defrag_difference

 

[taltamir]

to quote the conclusion of the article:

Don?t Waste Your Money or Time!

You shouldn?t break the bank for negligible performance gains

With all of the benchmarking completed, we find it rather suspicious that disk defragmentation did nothing to improve the performance of our machine. However, we must note that our test drive was not terribly fragmented to begin with due to Vista?s auto-defragger running on our test bed. Even the paid-for programs were unable to yield any positive gains?quite the opposite, in some instances.

We had high hopes for Diskeeper at first. Given the relatively high level of fragmentation it quoted compared to Vista?s built-in app, we assumed the program?s analysis routines were seeing fragmentation that Vista couldn?t. In turn, we expected Diskeeper to do a better job of moving files around and ultimately give us better benchmark numbers than the Vista client.

Funny they would say that about diskkeeper, considering microsoft licenses their defragmenting tool from diskkeeper (it is a stripped down version of their commercial tool). So why does it "find" more defragmentation? yet ends up with the same results?


Another thing about defragging...
1. You waste electricity doing all that work on the drive and CPU to defrag.
2. You waste lifespan of components, especially hard drives.
3. You waste time, because the defragging tools usually wanna run on a schedule, so you will often have system slowdowns while the defrag tool is defragging in the background... and for what? negligible performance gains?

 

[ChAoTiCpInOy]

I had to defrag my computer to open up my partition so I could partition my boot camp partition.

 

[spinejam]

Originally posted by: NXIL
http://www.maximumpc.com/artic...disk_defrag_difference

tx for the linky! :thumbsup:

 

[taltamir]

open a partition? what does that even mean? you mean mount it?

 

[martensite]

I 'defrag' regularly, or atleast, my copy of Diskeeper '09 does. I haven't bothered with manual defrags since the automatic background defrag mode works fine. I prefer it that way, and my system runs nicely.

Originally posted by: NXIL
http://www.maximumpc.com/artic...disk_defrag_difference

Not 100% sure, but I think that test is not done right.

I am not too familiar with PCMark vantage, but I believe it benchmarks the HDD (predetermined series of clusters?) much like HDTune, and not the file system as such. File fragmentation affects the file system.

File system fragmentation (or absence of) should not have any bearing on the results of that test. So the data they have obtained is just statistical noise I think, especially on Vista which has a mind of it's own wrt HDD activity during and after boot.

 

[martensite]

Originally posted by: taltamir
considering microsoft licenses their defragmenting tool from diskkeeper (it is a stripped down version of their commercial tool).

Another thing about defragging...
1. You waste electricity doing all that work on the drive and CPU to defrag.
2. You waste lifespan of components, especially hard drives.
3. You waste time, because the defragging tools usually wanna run on a schedule, so you will often have system slowdowns while the defrag tool is defragging in the background... and for what? negligible performance gains?

Vista's defragger is 100% MS. I am pretty sure I read it on the MS website somewhere.

1. Nonsense.
2. Nonsense.
3. Not the latest ones including Vista's. They are invisible.

 

[Lorne]

There stupid, Test are total flaw and as anyone knows 30seconds added to start up is a long time and shut down isnt as important but still 10 seconds means alot for reboots.
One problem is they ghosted which also reproduces the same fragmented FAT.
They needed to use the MFG copy software, A File copy image will produce a clean MSF/FAT.
Ive hear but as rumor that Vista defrag leaves little to be desired, But Im sure somone could que in on that wether 3rd party might be better.

Defrag makes a difference but it can be minimal or great change.
As Spindle drives have changed and evolved over the years different set ups can produce considerable differences in speed and change from file fragmentation.

This even goes with any OS (Yes Nix also), Theres are changes that accure durring clean ups and defrag that alter the statistics of how a OS handles file access.
An eg. I can give is a 250Gig drive used for say 5 years without defrag, If benched then defraged then benched then imaged off formated and imaged back again, Benched will produce 3 different benchmarks, Why?, Well for one the Files are located in different spots on the spindle, The other is that the MFS (FAT or what ever the initials elude me at the moment of writing) has also changed and in size and fragmentation.
This is also more noticable in OS that any kind FAT over BAM as its a redundancy of look up tables that become fragmented.

Some drive access times are fast enough to compensate but in the end the more the heads have to move and over larger distances the slower things get.
When HDs speed surpass controller throughput will defragging make no difference, But by that time SSD would be in full use and no need to defrag there.

And defrag doesnt lower the life of spindle drives, There expectancy is wether its on or off nothing more, There may even be less wear as the heads are travaling around insted of wearing a groove in the neutral idle path.

I think I will reproduce there test, But Im going to have to leave part of it out as I defrag often because I can tell when even my page file has fragmented to much.
But I will try both File image and ghosting and post up my findings.

 

[Gillbot]

I never defrag.

 

[Modelworks]

I only defrag certain directories .
Things like game directories or programs I use often.
I use wincontig. It is free and doesn't install services or other files.
http://wincontig.mdtzone.it/en/index.htm

 

[Blain]

On my HTPC I'm always adding and deleting HUGE files.
I do defrag that PC once every week or two, depending on how much has been deleted.

 

[Nothinman]

do you guys defrag your hd? if so, do you notice any benefits or is it just psychological?

It's just about all placebo except in corner cases. If you're using Vista it's already doing background defrags so just let it do it's thing.

There stupid, Test are total flaw and as anyone knows 30seconds added to start up is a long time and shut down isnt as important but still 10 seconds means alot for reboots

Most benchmarks are flawed in one way or another, it's just the nature of the beast. There's no good way to automatically benchmark real world usage.

One problem is they ghosted which also reproduces the same fragmented FAT.

If it does, your imaging tool is broken.

They needed to use the MFG copy software, A File copy image will produce a clean MSF/FAT.

No, any imaging tool that understands the filesystem being used will give you a defragmented image.

Ive hear but as rumor that Vista defrag leaves little to be desired, But Im sure somone could que in on that wether 3rd party might be better.

Lots of people say that but no one has been able to produce numbers to back it up.

Defrag makes a difference but it can be minimal or great change.
As Spindle drives have changed and evolved over the years different set ups can produce considerable differences in speed and change from file fragmentation.

The difference is almost always unnoticable. There's very few cases where it would make a real difference.

This even goes with any OS (Yes Nix also), Theres are changes that accure durring clean ups and defrag that alter the statistics of how a OS handles file access.

But most unix filesystem drivers are smart about file allocation so there's very little fragmentation to begin with, not that it makes much of a difference.

An eg. I can give is a 250Gig drive used for say 5 years without defrag, If benched then defraged then benched then imaged off formated and imaged back again, Benched will produce 3 different benchmarks, Why?, Well for one the Files are located in different spots on the spindle, The other is that the MFS (FAT or what ever the initials elude me at the moment of writing) has also changed and in size and fragmentation.

The MFT (assuming NTFS) should be the same size in all 3 cases because it's size is based off of a percentage of the drive size. Sure you can manually force it to be a different size but we're assuming you didn't do that. As long as you're not doing anything else while running the benchmark the numbers of all 3 runes should be very similar.

This is also more noticable in OS that any kind FAT over BAM as its a redundancy of look up tables that become fragmented.

No, FAT or MFT fragmentation has little to no affect because they're so clustered together anyway.

I think I will reproduce there test, But Im going to have to leave part of it out as I defrag often because I can tell when even my page file has fragmented to much.
But I will try both File image and ghosting and post up my findings

Pagefile fragmentation is irrelevant because that file is accessed randomly anyway and almost always amongst a lot of other paging activity. Go ahead give it a try if you like but all you'll really find is that getting an accurate benchmark is very difficult.

 

[taltamir]

Just to clarify, I don't defrag, but I am ok with anti fragmentation policies... have your torrent program preallocate the space to avoid fragmentation. This is nice because they are "Free operations".

martensite... go get a kill-o-watt meter, plug it in, and check your idle power draw vs power draw defragging. post it here to disprove me.