Personal computing discussed

Moderators: renee, Dposcorp

 
Igor_Kavinski
Minister of Gerbil Affairs
Topic Author
Posts: 2077
Joined: Fri Dec 22, 2006 2:34 am

O&O Defrag 23 Pro

Thu Dec 19, 2019 11:25 am

https://blog.oo-software.com/en/new-oo- ... idcomplete

The new defragmentation method SOLID/COMPLETE: SSDs save file fragments on morememory cells than needed. The resource and hardware friendly SOLID/COMPLETE method defragments these file fragments so that only the number of memory cells required will be used for reading and writing in the future. SOLID/COMPLETE can be used on both SSDs as well as harddisks.


Can anyone verify if SSDs really do save data inefficiently by utilizing more memory cells than necessary?

Anyway, it's currently discounted: https://shop.oo-software.com/36/purl-flash2019Defrag
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 11:41 am

Sounds like marketing BS to me. A defrag tool running on the host does not have any knowledge of (or control over) the wear leveling or block allocation algorithms being used internally by the SSD firmware, other than what is exposed via the TRIM logic.

It can certainly make some educated guesses about the flash page size, and try to consolidate files into as few logical pages as possible, on the assumption that this will in turn map to the fewest possible number of distinct physical pages. But this is really no different than making sure the blocks in a file are stored contiguously and sequentially, which is what a defragger is supposed to do anyway.

Kind of reminds me of how back in the late '80s and early '90s, ads for headphones all had to tout that they were "digital-ready", even though that was completely irrelevant since the connection was still analog. Or how monitors had stickers on them that said "Windows Vista Ready" when Windows Vista came out. Uhh... if you just correctly implement the basic functionality everyone would expect from a product of this type, it'll automatically do this thing you're promoting as if it is some sort of revolutionary new feature...
Nostalgia isn't what it used to be.
 
Aranarth
Graphmaster Gerbil
Posts: 1435
Joined: Tue Jan 17, 2006 6:56 am
Location: Big Rapids, Mich. (Est Time Zone)
Contact:

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 11:53 am

As a side note...

ntfs does have a important limitation that ssd users need to be aware of in that there is a maximum number of fragments a file can be.

Sure it's an obscene number, (1.5 million) but I still recommend dragging an SSD once a year and then then running trim.
This will keep your ntfs file system healthy.

Dragging once a year should not hammer the longevity of the drive much.

The other side of this is that if a file is badly fragmented it will use more disk space than it should. This is more noticeable with small files that have lots of fragments.
What happens is each fragment might be smaller that your sector size leading to wasted space. You might find you save a couple hundred meg on a very large drive that is badly fragmented.
Main machine: Core I7 -2600K @ 4.0Ghz / 16 gig ram / Radeon RX 580 8gb / 500gb toshiba ssd / 5tb hd
Old machine: Core 2 quad Q6600 @ 3ghz / 8 gig ram / Radeon 7870 / 240 gb PNY ssd / 1tb HD
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 12:06 pm

I was under the (possibly mistaken) impression that unless you are using on-the-fly compression, NTFS always allocates space in 4K (or larger) chunks, and that this is also the smallest possible fragment size. So any time you have a file that isn't a multiple of 4K you waste some space, and (compression aside...) a single 4K block never contains fragments of more than one file, nor does it have any wasted space unless it is the tail (last) block of the file.
Nostalgia isn't what it used to be.
 
Igor_Kavinski
Minister of Gerbil Affairs
Topic Author
Posts: 2077
Joined: Fri Dec 22, 2006 2:34 am

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 12:16 pm

Found a cool article: http://codecapsule.com/2014/02/12/codin ... -contents/

So my question is: Does the filesystem and its fragmentation have an impact on the virtual file table maintained by the SSD controller? Ideally, the controller should be designed so that it is filesystem-agnostic. But maybe some controllers are better in this regard than others?
 
just brew it!
Administrator
Posts: 54500
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 12:21 pm

Igor_Kavinski wrote:
Found a cool article: http://codecapsule.com/2014/02/12/codin ... -contents/

So my question is: Does the filesystem and its fragmentation have an impact on the virtual file table maintained by the SSD controller? Ideally, the controller should be designed so that it is filesystem-agnostic. But maybe some controllers are better in this regard than others?

I have seen occasional claims that some memory cards try to optimize their wear leveling strategy based on knowledge of exFAT internals. But that (sort of) makes sense, since exFAT is the official industry-sanctioned format for memory cards.

For HDDs and SSDs, such format-specific optimizations would make little sense since the manufacturer doesn't know ahead of time how the drive will be formatted - the file system will be completely different depending on whether the device is used with a Windows, Mac, or Linux system (or in a RAID array, for that matter).
Nostalgia isn't what it used to be.
 
meerkt
Gerbil Jedi
Posts: 1754
Joined: Sun Aug 25, 2013 2:55 am

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 12:53 pm

Filesystem knowledge at the storage-device level sounds like it would be both a lot of work, and unreliable.

Aranarth, I doubt NTFS/FAT can allocate less than a cluster (last cluster can be not-full, though).
 
Waco
Maximum Gerbil
Posts: 4850
Joined: Tue Jan 20, 2009 4:14 pm
Location: Los Alamos, NM

Re: O&O Defrag 23 Pro

Thu Dec 19, 2019 1:11 pm

Igor_Kavinski wrote:
Does the filesystem and its fragmentation have an impact on the virtual file table maintained by the SSD controller? Ideally, the controller should be designed so that it is filesystem-agnostic. But maybe some controllers are better in this regard than others?

Does it have an impact? Yes, because the way a file is written absolutely changes how it is laid out in the FTL on the drive. No standard controller does anything other than optimize layouts of incoming blocks as nicely on the flash as possible. It does not have the concept of a file, so it can't do anything to lay out a file "optimally" if that file is written in a bad way (random 4K for example).

So short answer, no, you can't really do anything about it aside from using a filesystem that is flash-aware and tries to do this sort of optimization itself or write your files in a better matter (also probably not possible for most people).
Victory requires no explanation. Defeat allows none.

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On