Multiplatform! JKdefrag on Linux?

[Mitch 74]

Could you see if there is a way to make command-line version of JKDefrag work on Linux/BSD with ntfs-3G driver? That would be terrific.

[the Irf]

A computer is like air conditioning: it's a waste when you open windows.

Have you tried nLite ?
http://nliteos.com/
You can change windows in many ways with nLite, see the home site
C: is about 500-600MB for me with everything I need installed.

[jeroen]

Could you see if there is a way to make command-line version of JKDefrag work on Linux/BSD with ntfs-3G driver?

JkDefrag is based on the Microsoft defragmentation API. As far as I know that API has never been ported and is simply not available on Linux.

[Donn Edwards]

Every time I ask I'm told that Linux doesn't need a defrag program. Is this really true?

[jeroen]

Every time I ask I'm told that Linux doesn't need a defrag program. Is this really true?

That's what I keep hearing, but I find it very hard to believe. I wonder if it's a myth, maybe everybody is repeating the same thing over and over? I'd be very interested in hard measurement data, if somebody has it!

[ewieldra]

hahahahaha

Microsoft Said NT4 doesn't need a defragger .. because NTFS bla bla bla ..

SURE

LINUX IS SUPERIOR ...

SURE

Jeroen you are right .. it's a myth

[elbiatcho1]

Found an interesting explanation: Why doesn't linux need defrag.

[jeroen]

Found an interesting explanation: Why doesn't linux need defrag.

Thanks for the link, interesting! I have the following points.

1. As the article says, Linux also uses filesystems (such as FAT, for memory sticks and USB harddisks) that need defragmentation. Everybody agrees on that, yet there is no decent defragger for Linux.

2. The article is basically a thought experiment. There is no hard measurement data, such as how much fragmentation occurs and what effect it has on access times.

3. The article only deals with fragmentation, not a word about optimization.

By the way, there is a defragger available for Linux, but it's very old and nobody uses it. It's dangerous because safe defragging requires special support deep inside the filesystem drivers, and Linux simply doesn't have that. To put it in other words: I think Linux cannot be defragged because it cannot be defragged. Not because it doesn't need to be defragged.

[oneandoneis2]

Hi,

as the writer of the linked article, I just wanted to make a couple of points:

It's not true that Linux has no defrag software. Wikipedia will tell you about a few of them. The most recent effort I'm aware of is one called shake and is under current development. And there's always the last-resort method of backing all files up to a redundant storage device and then copying them back. Crude, but effective if really needed.

Having said that, it's also not true that Linux filesystems need defragging and just don't get it: I just ran a check on my somewhat-older-than-a-year Linux ext3 partition and fragmentation was at 1.5% - there are many other users with similar experiences. Several of them left comments to that effect on the same article, IIRC.

Of course, if you still can't believe that it's true, Linux is available for free so why not install it on your PC and test it for yourselves? Then you'll KNOW how necessary it is or isn't  

[jeroen]

It's not true that Linux has no defrag software. Wikipedia will tell you about a few of them.

Interesting article! I've taken a look. The very old and unused defragger that I mentioned before is there (e2defrag). The JFS and XFS filesystems are new to me, never even heard of them. And there is no defragger for ext3, the default filesystem of many Linux distributions. So most Linux users simply have no defragger and no optimizer. Compare that with the (incomplete) list of 11 proggies for Windows!

The most recent effort I'm aware of is one called shake

I can only agree with the very last message on the forum when I looked just now: "This looks interesting. I am looking forward to a stable version."

fragmentation was at 1.5%

I'm a bit sceptical about that number. What kind of files on that partition? And 1.5 percent of what, the total partition size? And I've seen several postings on the Shake forum from people reporting totally different experiences.

[oneandoneis2]

That's the only partition I've got for that particular install, so it's got all the system files and all my personal files - MP3s, office documents, movies, firefox's cache... What you'd expect on any PC, really.

The 1.5% refers to the percentage of all files on the partition that are fragmented.

There are various ways of using a filesystem that can lead to more fragmentation. Not having enough free space is always the real killer, of course, but I gather certain P2P applications (perhaps since they work by manipulating fragments of files?) can also cause issues. Plus Linux users can be using any one of a number of filesystems, each of which will of course react differently to the same situations.

My personal FS preference is ext3, which is about as mature as Linux filesystems get, and there's gobs of free space, so there would have to be something badly wrong for me to see any more fragmentation than I've got.

People who are using newer filesystems, or putting a lot more demands on their systems, may well experience fragmentation problems. An average user sticking with "tried and tested" is much less likely to. All I can say for sure is that my own experience, in years of using a variety of Linux distros on a variety of hardware, has always been the fragmentation has never exceeded a couple of percent or caused any performance issues for me. I've certainly never felt a need to hunt for and install a defragmenter.

Optimization is, of course, a completely different topic, as it deals with the ordering of files relative to each other rather than the contents of individual files spread over the platters. Since switching to a SATA hard drive, I can't say hard drive performance has ever bothered me enough to even care about optimization, tho, I have to admit. You may find this worth a read tho: http://www.kdedevelopers.org/node/2270

YMMV, as ever. Like I say, if you really want to know firsthand, order a disc from Ubuntu (they send them free) and find out for yourself! There's a certain amount of faith involved when your source of information is "what it says on the Internet"

[jeroen]

The 1.5% refers to the percentage of all files on the partition that are fragmented.

On Windows the percentage is usually calculated by counting the size of all the fragmented files, and comparing it with how much of the partition is in use. See the "Number of clusters in fragmented items" at the end of the JkDefrag logfile, calculated relative to the "Number of clusters in analyzed items." This ofcourse results in a far higher percentage. Imagine a big fat fragmented movie, in Windows it will show up as several percents, in Linux the percentage doesn't even change, it's only 1 file. I think if you make a visual representation of your Linux disk (a disk map) that you will be shocked to see how fragmented your disk really is.

Defraggers on Windows are usually commercial defraggers. The makers are selling a solution to a problem, so naturally they tend to exaggerate. In the Linux community on the other hand there is a mechanism where enthousiasts tend to exaggerate the qualities of the operating system. Windows users are hyped into thinking they have a problem, Linux users are lulled into thinking they have not.

NTFS and EXT3 perform fairly similarly, both are mature and both contain anti-fragmentation algorithms. I think fragmentation occurs at about the same level on both, and that a proper defragger/optimizer for EXT3 will result in the same benefits as it does for NTFS.

[oneandoneis2]

On Windows the percentage is usually calculated by counting the size of all the fragmented files, and comparing it with how much of the partition is in use. See the "Number of clusters in fragmented items" at the end of the JkDefrag logfile, calculated relative to the "Number of clusters in analyzed items." This ofcourse results in a far higher percentage.

Undoubtedly it could result in a different percentage - higher OR lower. But unless there are very large and highly-fragmented files involved, I can't see it making anywhere near enough difference to bring it up to Windows levels. And I just don't have that many big files...

I think fragmentation occurs at about the same level on both, and that a proper defragger/optimizer for EXT3 will result in the same benefits as it does for NTFS.

Then I for one encourage you to write a Linux fragmentation measurer and fixer to answer the question once and for all!

[jeroen]

Then I for one encourage you to write a Linux fragmentation measurer and fixer to answer the question once and for all!

Do it yourself dude. You are the one with a Linux computer, and you are the one making all kinds of strong claims about fragmentation not being a problem on Linux. I am challenging your conclusions because they seem to be based on nothing. I'll believe you when you can show hard measurement data. Sorry...

[oneandoneis2]

Me? Why would I want to write something to fix a problem that I don't have?

Makes no odds to me whether you believe fragmentation to be a problem with Linux or not. I only chipped in at all because I saw posts being made based on "I don't know, but it stands to reason..." and thought that some actual information might be helpful.

I've already given you my own data, if you really have an interest in "hard measurements" I'll say again that nothing you'll read in a forum can be considered terribly definitive, and your best bet is to simply install and use Linux for a while and see what happens.

If you don't have such an interest, feel free not to bother. It's not like I'm on commission  :wink: