Jump to content
CCleaner Community Forums

Defraggler Time Taken to Defrag


Recommended Posts

Having had Piriform Defraggler for many years with my XP computer, I find that with my new Windows 10/1903 computer Defraggler is a big NO-NO.

In my latest defrag of around 4,000 fragments in 380 files, Defraggler was taking longer than a Windows Cumulative update.

After 45 minutes with around 1,000 fragments left, I could feel myself growing older and killed it as being a time liability.

Auslogics and SmartDefrag complete a defrag in a few tolerable minutes.

With sadness I bid Avast Defraggler a permanent Goodbye !

 

 

 

Link to post
Share on other sites

"We will soon be starting to plan for a new release of Defraggler. Are there more things that you wish Defraggler would do for you?  Post your suggestions here for Defraggler improvements and new features."

NO COMMENT !

Link to post
Share on other sites
  • Moderators

So to analyse that -

What you are saying is that after Defraggler had done most of the work with the difficult/more fragmented files you stopped it, and then a different defragmenter did the rest of the easy/less fragmented files quickly.

The different defragger(s) didn't have to deal with the difficult/more fragmented files because Defraggler had already done them.

So it's hardly a basis for a meaningful comparison.
It's not just the total number of fragments, it's how much each particular file is fragmented that counts.
For example the ones that Defraggler did may have had hundreds of fragments for each file, whereas the ones left after you stopped it may only have had one or two fragments per file.
You would need to know how long the other defragmenters would have taken with the original, more fragmented, files to make a meaningful comparison.

Which is why it's always difficult to compare different defragmenters, it's very difficult to make sure that they are starting with exactly the same amount of fragmentation. (You would have to make an exact mirror copy of the fragmented disc with all the fragments intact before starting a comparison).

So usually all you can rely on is your own user perception of which is the best for you.

 

Link to post
Share on other sites
15 minutes ago, nukecad said:

So to analyse that -

What you are saying is that after Defraggler had done most of the work with the difficult/more fragmented files you stopped it, and then a different defragmenter did the rest of the easy/less fragmented files quickly.

The different defragger(s) didn't have to deal with the difficult/more fragmented files because Defraggler had already done them.

So it's hardly a basis for a meaningful comparison.
It's not just the total number of fragments, it's how much each particular file is fragmented that counts.
For example the ones that Defraggler did may have had hundreds of fragments for each file, whereas the ones left after you stopped it may only have had one or two fragments per file.
You would need to know how long the other defragmenters would have taken with the original, more fragmented, files to make a meaningful comparison.

 

Sorry Nuke but you have it all wrong.

I have only used Defraggler a few times since December last when I got my new PC.

I have consistently used Auslogics many, many times, even after a massive Windows Feature update from 1809 to 1903 when fragmentation was in abundance.

Auslogics also did the hard slog with difficult and large files and did it well within 10 minutes - every time.

Defraggler cannot even tackle routine defragging remotely like that - Defraggler always takes ages to complete a defrag - even the simplest.

Its no use voicing the virtues of Defraggler I have months of practical experience and simply do not have enough life left to continue with the program.

Link to post
Share on other sites
  • Moderators

I wasn't defending/promoting Defraggler, simply pointing out that comparing different defragmenters logically is an almost impossible task.

You don't know exactly what each one is defragmenting, apart from the starting conditions usually being different one may be more thorough than the next, one may be ignoring less fragmented files, or so on.

As I say above it's down to user perception which suits them the best,  when you find one you like stick with it.

Link to post
Share on other sites
2 hours ago, nukecad said:

I wasn't defending/promoting Defraggler, simply pointing out that comparing different defragmenters logically is an almost impossible task.

You don't know exactly what each one is defragmenting, apart from the starting conditions usually being different one may be more thorough than the next, one may be ignoring less fragmented files, or so on.

As I say above it's down to user perception which suits them the best,  when you find one you like stick with it.

Fair enough Nuke.

All I can add is that after a defrag with Auslogics, an analyse  check with Windows built-in Defragger shows 0% fragmentation.

Link to post
Share on other sites
  • Moderators

The algorithm differences is why Defraggler takes way too long in my opinion, but they can only tweak that if they listen to the many years worth of user complaints about it being too slow and release a new quicker version -- it might mean less of a thorough defrag though. The slow complaints are archived in the comments section on software download sites too.

Link to post
Share on other sites

Having sung the praises of Auslogics, I must advise of an aspect of this program which is so unacceptable that I dumped it.

Progressively over the years I have repeatedly noticed using Malwarebytes, that Auslogics implant a huge number of Pups into the users system.

Malwarebytes recently found 61 Auslogics Pups on my system. That was the deciding factor - I completely uninstalled the program.

 

Link to post
Share on other sites
  • Moderators

I use a very old portable version of it I got off Softpedia.com, I only use it on my WinXP system though because on Win10 that old version looks microscopic.

Edit:

In that old portable version Malwarebytes gives a false positive against it though, it wasn't all PUP'd up back then, and even the desktop shortcut will make MBAM angry.

Link to post
Share on other sites
  • 2 weeks later...
Quote

 

A quick defrag took 13 minutes.

Windows seem to calculate % fragmentation on number of files.
They gave fragmentation at 3%. Total number of files on PC is 295,000, they gave 7,900 fragmented.
7,900/295,000x100 = 2.7%

Defraggler seems to consider GB, they gave 13% fragmentation - 6.9 GB out of 53.9 GB used.
6.9/53.9x100 = 12.8%

After defrag, Defraggler gave 14 files @ 4.8 GB still fragmented with 48 fragments - saying it is still
9% fragmented ! 
4.8/53.9x100 = 8.9% !!!

IMO "number of files" is the correct basis on which to consider fragmentation, NOT GB.
I will adopt Windows defrag, using "files" as the basis of fragmentation and wave goodbye to GB comparisons.

To deal in file fragments and then measure the fragmentation in GB is moving the goal posts.

Link to post
Share on other sites
  • 11 months later...

I am doing a once-after-6-months defrag of a 512GB Samsung SSD, which does 1 to 3 GB/s sequential reads, and usually around >600MB/s sequential writes.

In an effort to understand the fragmentation, I started by defragging the largest and most fragmented files (top 100, etc)

Now that I am looking at 500x 1MB files with ~4 fragments each, I am seeing <1 MB/s throughput for the last hour.

I don't know why Defraggler doesn't use a scatter-gather with a live bitmap to read a pile of files into memory (32GB DRAM) in sector order (will help HDDs more), then write them back to the new locations, preferably sequentially behind each other.  With the bitmap you mark files that you will read, and skip (or re-strategize) a file that hits a write to a read-pending cluster. 

This can be done multi-threaded instead of single-threaded.  One thread reading and writing, one committing changes to the bitmap, another tracks writes to completed files that can be removed from the bitmap.  I have 16 threaded CPU at 4GHz, which can saturate SSD 2GB/s with 1 thread..

This should run at least 10 to 20x faster since it isn't throughput constrained.

First time I ran Defraggler was what seems like 10 years ago.  It works.  But has always been the slowest of the lot.

I would pay $100 for a version that was full speed.

Link to post
Share on other sites
  • Moderators

This is an old thread.

Please check the date of the last post on a thread before replying.

This old thread has now been locked.

Link to post
Share on other sites
Guest
This topic is now closed to further replies.
×
×
  • Create New...