What does defragmentation do?

[zoltan]

Most PC runs Windows, and therefor the HDD must be NTFS in order for Windows to operate properly/

If you are saying that most people are running on Linux, then you live in a dream land.

 

Some benchmark on a pro Linux... sorry I have to do like you... Lamelux.... web site, doesn't make your argument valid.

Also, benchmark performance is meaningless. Real world performance is what maters. Not theories.

 

1- TRIM keeps SSD healthy, not make things faster.

2- You do not need AHCI for TRIM in Windows. You are miss informed. However, people enable AHCI for Native Command Queue (NCQ) support for SSDs and mostly HDDs. NCQ is a AHCI feature.

3- TRIM not working on an encrypted drive is a TRIM limitation not Windows.

4- Its the SSD controller that fragment data on purpose, not Windows.

 

- I said that the most amount of storage space (datacenters, cloud infrastructure) on earth is governed by UNIX/Linux filesystems, you said that most "computers" use Windows, which is not true, Windows is the prevalent consumer PC platform, which is only a very small part of all computers. Every day since a couple of years, there are 1.5 million new activations of linuxbased RISC devices, all serious network infrastructure and high performance computing platform runs UNIX/Linux. There are more people that use linux devices than there are people that use windows devices (parking meters, heart monitors, smart TV's, phones, cars, navigations systems, printers, media centers, etc... all linux based, there isn't even competition with linux if you take all computing devices into account, the desktop PC is the only exception, but in comparison to the number of linux devices that's only a small amount of "computers". In short, I claim that both in terms of number of "computers" and in terms of number of Petabytes, linux/unix filesystems govern a lot more devices and amount of storage than NTFS.

 

- So you don't accept the benchmark because you don't like the site it's published on, of because it shows that ext4 is over 25% faster than NTFS? If you think it's not true, please feel free to post a link to a benchmark that concludes the opposite. And if you think real world performance is different, maybe try a linux device some time, the experience should speak for itself.

 

- ad 1: trim is a wear-leveling mechanism, a functionality that flags stored data as no longer necessary. I said literally "SSD's rely on trim to stay healthy", and you reply to that "TRIM keeps SSD healthy, not make things faster", so I think you've not read my post correctly;

- ad 2: Windows TRIM only works for AHCI SSD's and TRIM is not natively supported by the Windows kernel (in contrast to the linux kernel) and requires drivers, just like I said. Source: https://social.technet.microsoft.com/Forums/en-US/w7itprohardware/thread/2c7cbc60-5878-4ff1-b013-cf482b499c05

Other sources are also available, I have not found any source that states the opposite. Please feel free to post your source.

- ad 3: TRIM works on AHCI mode SSD's with a driver provided by the SSD manufacturer in Windoze (whereas it doesn't need a driver or a specific mode in linux, it works with most filesystems there), but TRIM in Windoze exposes the encryption because it's done by a supplementary piece of software (even if the SSD does it's own encryption, because in Windoze it will need the driver to do that, it's not supported by the kernel like in linux), which means that it's a typical Windoze vulnerability that cannot be solved. If you want to safely use an encrypted SSD with wear-leveling, you can't use Windoze. Using userspace software application to encrypt wear-leveling enabled SSD's in linux has the exact same issue as Windoze, but in linux, the native encryption functions are there, so you don't have to use the userspace applications, so there is no security problem. Source: http://www.truecrypt.org/docs/?s=trim-operation. Please feel free to post the source you've based yourself on to discard my argumentation.

- ad 4: Please expand on this, as I don't understand what you mean, and please post your source. Do you mean SSD controllers cause fragmentation, not Windoze with NTFS? I think it's considered best practice not to defragment SSD's, as there is no gain in defragmenting it, I think defragmentation is necessary for HDD's formatted with fragmenting filesystems like NTFS only.

[GoodBytes]

1- You are changing the path of the topic of conversation. We are talking about PC.. not phones, not data-centers, not OSs in microwaves. And don't start talking about how PC can be a server and microwaves... and all that crap, by playing with the wording of definitions. It's a forum on PC's and PC is what is assumed, unless specified otherwise. You know this perfectly well, but you hate to be proven wrong, so you switch things, to try and make others look like they don't know what they are talking about. This is not particularly respectful.

 

2- The reason why data centers uses Linux and Ext4, is that: For 1- Ext4 is fully supported by Linux (obviously), not NTFS. Therefore, Ext4 is, obviously, a better choice.  Also, Linux is best at data centers over Windows Server. Every OS has their strength and weaknesses.

 

3- On TRIM, I don't care that some guy said on the support forum of Microsoft, that AHCI needs to be enabled. Stop living in a perfect world, and understand that. NORMALLY, TRIM is apart from AHCI. You don't need AHCI.

However, they are some SATA Controller, that when you set them to compatibility mode or Legacy, or IDE, or whatever the BIOS wants to call it, it makes the SATA Controller ignores TRIM command, and I even encounter some that completely emulated IDE, down to limit its speed (my old computer did this, however my current computer does not).

 

4- I don't need to state source for common knowledge. Do you want sources, interviews, and bring an engineer at your place, to tell you that cars have wheels?

[zoltan]

1- You are changing the path of the topic of conversation. We are talking about PC.. not phones, not data-centers, not OSs in microwaves. And don't start talking about how PC can be a server and microwaves... and all that crap, by playing with the wording of definitions. It's a forum on PC's and PC is what is assumed, unless specified otherwise. You know this perfectly well, but you hate to be proven wrong, so you switch things, to try and make others look like they don't know what they are talking about. This is not particularly respectful.

 

2- The reason why data centers uses Linux and Ext4, is that: For 1- Ext4 is fully supported by Linux (obviously), not NTFS. Therefore, Ext4 is, obviously, a better choice.  Also, Linux is best at data centers over Windows Server. Every OS has their strength and weaknesses.

 

3- On TRIM, I don't care that some guy said on the support forum of Microsoft, that AHCI needs to be enabled. Stop living in a perfect world, and understand that. NORMALLY, TRIM is apart from AHCI. You don't need AHCI.

However, they are some SATA Controller, that when you set them to compatibility mode or Legacy, or IDE, or whatever the BIOS wants to call it, it makes the SATA Controller ignores TRIM command, and I even encounter some that completely emulated IDE, down to limit its speed (my old computer did this, however my current computer does not).

 

4- I don't need to state source for common knowledge. Do you want sources, interviews, and bring an engineer at your place, to tell you that cars have wheels?

 

1- Excuse me, but I have not been proven wrong, you have proven nothing, you have only formulated ad hominem arguements against me. You said that 90+% of all "computers" (sic) use NTFS, which is not true, you didn't say PC's, and you didn't say consumer PC's. How can you call me disrespectful if you don't stay with the conversational subject but only use ad hominem unfunded meta-arguements?

 

2- Data centers will certainly never use Ext4, ext4 is for small systems, data center use fielsystems like GFS or XFS etc... Contrary to what you claim, NTFS is fully supported by linux through NTFS-3G, and supports FUSE (https://en.wikipedia.org/wiki/NTFS-3G). Please post your source if you claim the opposite.

 

3-4- I note that you don't care about the Microsoft people that post Microsoft specs on the Microsoft support forum, and I note that you have not posted a source to claim the contrary to what I have brought forward to dispute your earlier statements by which you have vigorously rejected my first statements in this thread. Please post a source to corroborate your statements out of respect for the forum discussion, if it is common knowledge as you claim it is, that should not be a problem for you. I'm not impressed by mockery, it may be commonly considered a sign of weakness and insecurity.

[GoodBytes]

The day you'll learn to stop playing with people words and yours to fit your needs to win an argument.. let alone know how to make one, I'll stop ignoring you.

End of conversation.

[Jack.EXE]

I'll put it as simple as I can, so here I go.

 

 

Imagine you have a big box of cups, you know, like plastic Solo cups, they are not stacked or anything, they're just thrown in there all willy nilly. This is sort of how a hard drive is. Now, say you go and stack all of the cups up into neat, even stacks. the whole pile of cups now fits into a space half the size. This is what happens when you defragment it.

[Huntsman]

If it is a fixed RPM there is not a faster region. The middle just wold hold less information than the outer region. The whole disk is moving at the same rpm no faster or slower region. I didn't read the article I just know from experience and common sense. The inner region would hold less information while the outer would hold more. But in the centre the head of the head would have to move less distance if the information was all centred. That might not be how it is but there is no faster region because (inner vs outer) only the head moving would make a difference.

 

uh.. I disagree. This is simple maths. Yes, the outer sector holds more info than the inner one, thats why when the platter spins at a constant speed, the header would go past the outer sector faster than it does the inner. It would cover more ground in the same amount of time compared to when the header is at the inner sector.

 

Another way to look at this is the mathematical way. We all know that mathematical expression for arc is L=theta*2*pi*r, where r is radius, pi is the ratio between the radius and circumference of a circle, theta is the internal angle between two radius and L is the lengh of arc.

 

 

Assuming same speed, the platter would have spun for same amount of theta. Thus the theta is a constant, let's assume it spun for 30 degrees. Also assuming the inner sector has a radius of 1 and outer sector has a radius of 3. The distance travelled over the platter by the header is L, which is the length of arc.

 

For inner sector,

L=30*2*pi*1

  =60pi

 

For outer sector,

L=30*2*pi*3

  =180pi

 

Don't mind the pi, is 180 of the outer sector bigger than 60 of the inner? Yes it is. So this kinda proves that data stored at the outer sector can be read faster than the ones in the inner sector since at the same RPM at the same amount of time, the header can cover more ground. Also, the access times (assuming all data at same radius from center) is lower. If the data is at different radius, that is when NCQ comes in.

 

Hope this isn't a wall of text that nobody understands...

[SkilledRebuilds]

Run any decent defrag program, and what it does becomes pretty self explanatory.

^

So the one inbuilt into windows doesnt count.

[GoodBytes]

Built-in Windows one is decent. It's just doesn't give you status update, and current state detail info of the drive. And not the fastest in the world, but definitely not slow. It's just average.

The work is does is fine.

 

BUT, here the thing that I forgot to mention. If you have Windows Vista or newer, and you use another defrag tool beside the built-in one, they will break  the system restore points it creates. So you won't be able to go back in time your system, until a restore point is made after the defrag process. The reason for this, is that the data stored in the System Restore points are just difference bits between the original files. This is amazing technique to save space and have a backup, however a defrag software, beside Windows, doesn't know what it is, and thinks it bit of data associated with files, and will try to fit it and fail, and put it somewhere else, or simply move it to another area of the disk, and now Windows doesn't know where it is. I am no experts, but I think this problem could have been solved if Microsoft made an API system, to see where the system restore points location is, so that defrag software doesn't touch it.

[Huntsman]

Built-in Windows one is decent. It's just doesn't give you status update, and current state detail info of the drive. And not the fastest in the world, but definitely not slow. It's just average.

The work is does is fine.

 

BUT, here the thing that I forgot to mention. If you have Windows Vista or newer, and you use another defrag tool beside the built-in one, they will break  the system restore points it creates. So you won't be able to go back in time your system, until a restore point is made after the defrag process. The reason for this, is that the data stored in the System Restore points are just difference bits between the original files. This is amazing technique to save space and have a backup, however a defrag software, beside Windows, doesn't know what it is, and thinks it bit of data associated with files, and will try to fit it and fail, and put it somewhere else, or simply move it to another area of the disk, and now Windows doesn't know where it is. I am no experts, but I think this problem could have been solved if Microsoft made an API system, to see where the system restore points location is, so that defrag software doesn't touch it.

 

oh.. that explains it.. lol i was racking my brain and asking so much forums back then and couldn't get an answer to this. Thanks!

 

p.s. microsoft releases good os, then bad os, then good, then bad.. so windows 8 should be the bad os LOL

[GoodBytes]

oh.. that explains it.. lol i was racking my brain and asking so much forums back then and couldn't get an answer to this. Thanks!

 

p.s. microsoft releases good os, then bad os, then good, then bad.. so windows 8 should be the bad os LOL

LOL yea... but sadly it's not true. It's not like the old StarTrek movies. :)

If you assume that XP is good. Windows 2000 was good as well, hence breaking the link. Also, where do you place NT 3, NT4 in the mix? And is Win95 revolutionary or bad?

Sorry for breaking the joke. But good one! As you can say that the initial perception of the OS is good and bad... maybe (but that is still questionable)? :)

[Huntsman]

Maybe there has to be a era limitation. Perhaps this phenomena happens only after XP? XP being good, Vista suck(not even bad, it suck), 7 great, 8... well, you can fill in the blanks xD

 

p.s. im a 90s kid, so my computing experience only starts with windows 98 and even so I was too young at the time to judge anything