Fragmentation Puzzle: How many ways can it be measured?

Hairy Scot

Power User
VIP Member
Power User
Messages
220
While playing with various defrag solutions I noticed a glaring difference in the fragmentation percentages reported by the analyse phase of each product.
On my 100GB Windows C drive:-
1. The built-in Windows 8.1 optimiser showed 0%
2. Defraggler showed 9%
3. Powersuite showed 31%
4. Perfect Disk showed File fragmentation: 0.1% Free Space fragmentation: 7.9%

:shock:
 

My Computer

System One

  • OS
    Windows 8.1 Pro X64
    Computer type
    PC/Desktop
    System Manufacturer/Model
    Gigabyte Technology Co., Ltd. GA-B75M-D2V
    CPU
    Intel(R) Core(TM) i5-3470 CPU @ 3.20GHz, 3601 Mhz, 4 Core(s)
    Motherboard
    Gigabyte Technology Co., Ltd. GA-B75M-D2V
    Memory
    8 GB DDR3 1600MHZ
    Graphics Card(s)
    GE Force GT 610
    Sound Card
    Onboard Realtek High Def
    Monitor(s) Displays
    Dell U2913WMP
    Screen Resolution
    2560x1080
    Hard Drives
    Western Digital WD10EZEX-00RKKA0 1TB
    Seagate ST2000DL003-9VT166 2TB
    Seagate ST31000528AS 1TB
Hi again Hairy Scot,

I do not believe there is any puzzle here because the determination of "how much" fragmentation is present is dependent on the parameters of the program being used. Some defrag programs may want to pack everything up to the front of the drive and establish consecutive sectors for each program; whereas, others may just want to move things around (as required) to make each program's sectors consecutive. Now, in my hypothetical example, I would have to believe that the one trying to pack everything up to the front with consecutive sectors would indicate much more defragmentation than the other.

Also, some defrag programs may include a "most frequently used programs" parameter and others may not even give that parameter any consideration.
 

My Computer

System One

  • OS
    Windows 8.1
    Computer type
    PC/Desktop
    CPU
    Intel G2020
    Motherboard
    ASRock B75M-DGS R2.0
    Memory
    8GBs @ 1333 MHz
    Hard Drives
    Samsung 840 EVO
    PSU
    400w
    Internet Speed
    57/11
Fragmentation Puzzle: How many ways can it be measured?
As many as the people making the calculation.

The calculation of the degree of fragmentation is not a simple thing. To do it at all a software product has to follow a set of rules considering many factors and make a number of essentially arbitrary decisions. There is no consensus on how to do this so each product has a different set of rules. It is to be expected that the results will be different, sometimes to a considerable degree.

I have never found those fragmentation numbers to be very useful. Comparing the results of different products is virtually meaningless.
 

My Computer

System One

  • OS
    Windows 7
    Computer type
    PC/Desktop
Fragmentation Puzzle: How many ways can it be measured?
As many as the people making the calculation.

The calculation of the degree of fragmentation is not a simple thing. To do it at all a software product has to follow a set of rules considering many factors and make a number of essentially arbitrary decisions. There is no consensus on how to do this so each product has a different set of rules. It is to be expected that the results will be different, sometimes to a considerable degree.

I have never found those fragmentation numbers to be very useful. Comparing the results of different products is virtually meaningless.

:ditto: Well stated!
 

My Computer

System One

  • OS
    Windows 8.1
    Computer type
    PC/Desktop
    CPU
    Intel G2020
    Motherboard
    ASRock B75M-DGS R2.0
    Memory
    8GBs @ 1333 MHz
    Hard Drives
    Samsung 840 EVO
    PSU
    400w
    Internet Speed
    57/11
Back
Top