Defrag or not to Defrag that is the question?

I read an article on another forum and I don't remember where, but it was about 2 months ago and the person said that he "never" defrags his HDD because the benefit is out weighed by the abnormal wear on the drive. Can this possibly be true? Until then I would defrag at least once a week, now I'm not sure.

Thanks,
HelpFromFrance

 

its not rele true. it does ware and tear your disk but its very little and will have a minute effect on disk's life span.

in fact, having fragmented files can also causes ware and tear because the disk heads have to move around to find all the parts to a file.

again, it is an insignificant issue and hard disks are quite reliable and long lasting anyways.

 

WSFuser,

Thanks I will go back to defraging my HDD. So is once a week too often about right or what?

Thanks again
HelpFromFrance

 

its up to you.

i usually defrag before making a system image and after installing a game such as Half-Life 2 or UT2004. other than that, i rarely defrag.

 

Thanks for the advice. I will do it just before imaging.

HelpFromFrance

 

No.

There are a number of individuals who make such statements.
In some cases, thet just need to be educated, in other cases, they may be auditioning for a sequel to the movie "Clueless".

Get a good defrag program, such as Perfect Disk, and you
will not need to defrag as often,

 

How often depends on individual needs.

Perfect Disk comes with a Trend Analysus option that can run, say, once per day. Raxco provides, e.g. an Access database that facilitates analyzing the results. See the white papers at:

http://www.raxco.com/products/perfectdisk2k/wp.cfm

 

I usually defrag before imaging (PerfectDisk 7). I just do it without any enthousiasm.

 

I use Diskeeper and have for awhile. I have it set to Defrag only when needed. I check it every now and then to see how it's doing and my system remains in good health.

 

PerfectDisk 7, automated runs offline first and then online once a week. I neither analyze nor agonize over the true need for it. I just tweak the Rarely and Frequently Modified ages to roughly balance the three populations, set it to run and forget it.

 

Thanks to all,

HelpFromFrance

 

I just use the default settings in Perfect Disk, and it does amazingly well.

 

Agreed.

No !

Makes no difference as disc access is random not sequential (unless your doing something like on the fly movie editting).

NTFS only requests parts of a file at a time and different files.
There is also a lot of page file activity which causes a lot of head jumping anyhow.

If you run sysinternals filemon you will see that disc access is very random.

Also, when you defrag your drive you are only defragging LOGICALLY.

You do not know on the drive exactly which platter on the drive the file OR part of the file sits - its PHYSICAL location.

Heres a snip of my filemon:
Accesses are not logically sequential (we dont know if they are physically).

Note also the small amounts of data requested (the length) very small.

Code:

Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Excl: Yes Offset: 1073741824 Length: 1	
24603	11:14:02	firefox.exe:952	FASTIO_LOCK	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Excl: No Offset: 1073741826 Length: 510	
24604	11:14:02	firefox.exe:952	FASTIO_UNLOCK	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 1073741824 Length: 1	
24605	11:14:02	firefox.exe:952	FSCTL_IS_VOLUME_MOUNTED	C:\Program Files\Mozilla Firefox	SUCCESS		
24606	11:14:02	firefox.exe:952	FASTIO_QUERY_OPEN	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db-journal	NOT FOUND	Attributes: Error	
24607	11:14:02	firefox.exe:952	FASTIO_QUERY_STANDARD_INFO	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Length: 267264	
24608	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Read: Offset: 0 Length: 1024	
24609	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 0 Length: 1024	
24610	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Read: Offset: 120832 Length: 1024	
24611	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 120832 Length: 1024	
24612	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Read: Offset: 121856 Length: 1024	
24613	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 121856 Length: 1024	
24614	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Read: Offset: 160768 Length: 1024	
24615	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 160768 Length: 1024	
24616	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Read: Offset: 144384 Length: 1024	
24617	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 144384 Length: 1024	
24618	11:14:02	firefox.exe:952	FASTIO_UNLOCK	C:\Documents and Settings\rhodesn\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\googlesafebrowsing.db	SUCCESS	Offset: 1073741826 Length: 510	
24619	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Read: Offset: 7424 Length: 256	
24620	11:14:02	firefox.exe:952	FASTIO_READ	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Offset: 7424 Length: 256	
24621	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Write: Offset: 7424 Length: 256	
24622	11:14:02	firefox.exe:952	FASTIO_WRITE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Offset: 7424 Length: 256	
24623	11:14:02	System:8	IRP_MJ_CLOSE 	D:\Inetpub\wwwroot\DotNet\WebApps\Delivery_Management\Images\LeftHeadAlt.png	SUCCESS		
24624	11:14:02	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1066673 Length: 335	
24625	11:14:02	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1066673 Length: 335	
24626	11:14:02	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1067008 Length: 15548	
24627	11:14:02	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1067008 Length: 15548	
24628	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Write: Offset: 225792 Length: 512	
24629	11:14:02	firefox.exe:952	FASTIO_WRITE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Offset: 225792 Length: 512	
24630	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Write: Offset: 105216 Length: 256	
24631	11:14:02	firefox.exe:952	FASTIO_WRITE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Offset: 105216 Length: 256	
24632	11:14:02	firefox.exe:952	FASTIO_CHECK_IF_POSSIBLE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Write: Offset: 197120 Length: 512	
24633	11:14:02	firefox.exe:952	FASTIO_WRITE	C:\Documents and Settings\rhodesn\Local Settings\Application Data\Mozilla\Firefox\Profiles\a4omuuod.default\Cache\_CACHE_001_	SUCCESS	Offset: 197120 Length: 512	
24634	11:14:02	webtool.exe:1372	FASTIO_CHECK_IF_POSSIBLE	C:\Program Files\Microsoft Web Application Stress Tool\WAS.mdb	SUCCESS	Read: Offset: 0 Length: 2048	
24635	11:14:02	webtool.exe:1372	FASTIO_READ	C:\Program Files\Microsoft Web Application Stress Tool\WAS.mdb	SUCCESS	Offset: 0 Length: 2048	
24636	11:14:02	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 6168576 Length: 4096	
24637	11:14:02	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 12288 Length: 4096	
24638	11:14:02	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 8192 Length: 4096	
24639	11:14:02	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 24576 Length: 4096	
24640	11:14:02	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1082556 Length: 836	
24641	11:14:02	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1082556 Length: 836	
24642	11:14:02	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1083392 Length: 15046	
24643	11:14:02	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1083392 Length: 15046	
24644	11:14:03	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 6172672 Length: 49152	
24645	11:14:03	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 8192 Length: 4096	
24646	11:14:03	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 0 Length: 4096	
24647	11:14:03	System:8	IRP_MJ_WRITE*	D:	SUCCESS	Offset: 26230784 Length: 61440	
24648	11:14:03	System:8	IRP_MJ_WRITE*	D:	SUCCESS	Offset: 8192 Length: 4096	
24649	11:14:03	System:8	IRP_MJ_WRITE*	D:	SUCCESS	Offset: 0 Length: 4096	
24650	11:14:03	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1098438 Length: 1338	
24651	11:14:03	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1098438 Length: 1338	
24652	11:14:03	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1099776 Length: 14545	
24653	11:14:03	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1099776 Length: 14545	
24654	11:14:03	WINLOGON.EXE:188	IRP_MJ_CREATE 	\Device\LanmanRedirector	SUCCESS	Options: Open  Access: All	
24655	11:14:03	WINLOGON.EXE:188	FASTIO_DEVICE_CONTROL	\Device\LanmanRedirector	FAILURE	IOCTL: 0x140FCF	
24656	11:14:03	WINLOGON.EXE:188	IRP_MJ_DEVICE_CONTROL	\Device\LanmanRedirector	SUCCESS	IOCTL: 0x140FCF	
24657	11:14:03	WINLOGON.EXE:188	IRP_MJ_CLEANUP	\Device\LanmanRedirector	SUCCESS		
24658	11:14:03	WINLOGON.EXE:188	IRP_MJ_CLOSE 	\Device\LanmanRedirector	SUCCESS		
24659	11:14:03	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 0 Length: 24576	
24660	11:14:03	System:8	IRP_MJ_WRITE*	C:	SUCCESS	Offset: 12288 Length: 4096	
24661	11:14:03	System:8	IRP_MJ_SET_INFORMATION*	C:	SUCCESS	Length: 20639	
24662	11:14:03	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1114321 Length: 1839	
24663	11:14:03	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1114321 Length: 1839	
24664	11:14:03	winamp.exe:3004	FASTIO_CHECK_IF_POSSIBLE	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Read: Offset: 1116160 Length: 14043	
24665	11:14:03	winamp.exe:3004	FASTIO_READ	D:\Music\Iron Maiden - Best of the Beast (2CD).256k\Best of the Beast (CD2)\13-Strange World.mp3	SUCCESS	Offset: 1116160 Length: 14043	
24666	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 46125056 Length: 4096	
24667	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 128028672 Length: 4096	
24668	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 154755072 Length: 4096	
24669	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 63967232 Length: 4096	
24670	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 8015872 Length: 4096	
24671	11:14:03	WINLOGON.EXE:188	IRP_MJ_READ*	C:\pagefile.sys	SUCCESS	Offset: 40255488 Length: 4096	

 

Not so.

Defragging moves the file clusters so that they are contiguous for each file.

 

No, that is only logically contiguous.

For starters defragging in NTFS does not bypass the drivers, therefore does NOT access any physical mappings, only the logical mappings provided by the driver.

Then looking at a phyiscal level - how do you (the OS, the defragger) know which platters/tracks the files going to sit on.

The file could well be logically contigous (according to a block map), but could be physically split across tracks and platters.

the drive mapped track by track one platter at a time, or track by track for each platter.

Are the tracks on each platter mapped as one big track or seperate tracks and are they counted per platter or a track at a time ?

Most harddrives keep a spare chunk of sectors at one end or the other to remap when an existing one comes erronous... yet the OS will not know.

What about defragging a 3 hdd's that are stripped together as one big RAID array, how do you know that file is phyisically contiguous or split across one or more drives ?

 

Defragging moves the clusters so that they are contiguous in the sector address space.

How the sector address space is mapped by a drive is transparent to software.

For example, ISO/IEC 13346-2 contains the following:

And ISO/IEC 13346-1 contains the following:

 

Hello,
That's problematic.
What if you partition a drive with data on it? Even if it shows 50% free space, by what you say, you could be actually splitting clusters between partitions. Does that mean that even though logically you have 2 partitions, data could be read off the same physical area on the hdd? Does that mean that damage could cause 2 partitions to die simultaneously, even though they are logically separated? And how is partitioning done - I'm taking about existing, non-empty partitions. Are only logical addresses moved or are clusters physically copied from sector to sector to create physical separation between partitions?
Mrk

 

Gee guys, I did not mean to start a war

HelpFromFrance

 

Those are specifications for CD and DVD R and RW and have nothing to do with hard drive allocation or mapping.

 

Hi nickR,

OK, now that I'm over the shock and awe part, surely the logical mapping to physical sectors is logical, i.e., in some pattern that has a geometric regularity.

Even though sequentially addressed sectors could be on different platters, wouldn't at least the majority of these sequences be at the same and adjacent head position ?

Surely it's not as bad as files parts scattered randomly around the physical hard drive. Can you clarify it for me ?

 

Sectors are physically grouped using logic, the thing is it is different to the logically grouping of sectors at the OS level. Majority of the data will be organised, its just modern drives do have variable sectors per track for example, which makes LBA addressing unpredictable.

The logical to physical mapping also becomes inconcistant if defective sectors are swapped out.


Heres something I dug up from Microsoft:

http://www.microsoft.com/technet/scriptcenter/guide/sas_fsd_attu.mspx?mfr=true

 

No they are not scattered, they are logically laid out and wont badly fragment, id guess we are talking a tiny percentage of inconcistancy.

I am just trying to explain that logical maps are not the same as physical maps.

"When the 8.4 GB limit of the Int13h interface was reached in 1998-1999, it became impossible to express the geometry of large hard disks using cylinder, head and sector numbers, regardless of whether translated or not, while remaining below the Int13h limits of 1,024 cylinders, 256 heads and 63 sectors. This is one of the reasons that today's hard drives no longer indicate their classical geometry."

http://www.dewassoc.com/kbase/hard_drives/lba.htm

 

Defragging physically moves clusters.

The idea is to layout the file to minimize the need for disk seeks to get the file content.

THe underlyng recording standard/spec may indeed physically layout lolgical sectors non-contiguously, but as far as software is concerned sectors are physicallly contiguous.

Perfect disk will show where a file lives.

So, could check where the clusters are before and after defragmentation.

 

THe specifications of ISO/IEC !3346 are not device dependent.

It describes the way ALL file systems work in terms of volume space, etc.

Defragging relocates the clusters to minimize needed seek time.
Discussion of lower level details is pointless.

 

I did'nt say it was device dependant but it is designed for media, not fixed devices.

Heres a correct context of usage: http://osta.org/technology/di.htm

No !

It specifies methods of physically writting and rewritting media, it is itself a file system.

But head seeking is a low level operation.


Fujitsu for example actually state on their full specs that they vary the sectors per track eg 798/441 max min for the MPG 3204AT. A file which will fit on one track at one end of the drive will have to be physically fragmented (but only need to be split into 2 fragments) to fit at the other end of the drive.

Lets not forget if a file COULD fit on a track and there is already defragmented data on that track, that file will have to be split into 2 (or more) phyisical fragments to be logically defragmented.

It will incur a seek time penatly - but only a very small one.

In the grand scheme of things this does'tn matter one bit, due to hardware caching done on the drive (for read/writes) and OS caching and also the random access behavior of a multitasking OS as I mentioned above.

Which is also why only heavy fragmentation effects normal usage performance.