Diskeeper 10

[Mastertech]

A little background on NTFS and how it works. When an NTFS drive is formatted, it creates the $MFT and it contains a fixed number of records that can be used. As files are created, then these records are used. Eventually, the pre-allocated number of records is filled up and the $MFT needs to "grow" - allocating another chunk of file records. Immediately adjacent to the $MFT is created what is called the MFT Reserved Zone. By default, it is 12.5% of the drive and goes from the first record of the $MFT to the first non-free cluster after the last MFT record. If you have a 100GB drive, then the Reserved Zone is going to be 12.5GB - pretty large if you think about it. If you look in Windows Explorer/Properties on a drive, the free space shown is the total of both INSIDE and OUTSIDE of the Reserved Zone.

The MFT Reserved Zone is created specifically to allow the $MFT to "grow" in a contiguous fashion. When the $MFT fills up, it allocates the next chunk from the free space located in the MFT Reserved Zone. NTFS will avoid putting files inside of the Reserved Zone unless you get into a low free space condition. It is not necessary to artifically pre-allocate additional space for the $MFT in order to keep it from growing fragmented.

You don't understand Diskeeper takes into account the reserved space and only recommends extending the reserved space when it is near capacity.

Since NTFS takes care of this automatically, there is no need to artificially extend the $MFT. However, since most people don't have this level of knowledge of how NTFS works, when they hear about this "feature" in Diskeeper, they think it must be a good thing and worth paying extra money for :)

Now you are making things up. Diskeeper is not extending the MFT and disregarding the reserved space. It is taking into account the current MFT size and the amount of left over MFT reserved space. When the reserved space gets near capacity it recommends extending it to prevent the MFT from becoming fragmented. This has only a positive effect on file fragmentation because like you said when the disk gets near capacity any MFT reserved space can be used by Windows for regular files. Talk about misleading! NTFS does not increase the reserved space contigously. It increases it by fragmenting to the nearest area of free space on the disk.

The MFT configuration tool helps pre-extend the MFT in a contiguous manner, so future growth of the MFT will not result in fragmented extensions of the file.

[ghayes]

Now you are making things up. Diskeeper is not extending the MFT and disregarding the reserved space.

I'm not making things up.

Perhaps you didn't see this in the Diskeeper help file when you looking for something to post:

"Approximately one spare file record is needed for each file that will occupy the volume in the future. The number of file records to add is determined in one of two ways:

Frag Shield recommends the size increase based on the estimated number of files that could occupy the volume. This estimate is based on the current average file size and amount of available free space.

You can enter in how many files you estimate will potentially occupy the volume."

and

"Note: Once the MFT is extended, it cannot be reduced in size without reformatting the volume."

They ARE artificially extending the $MFT. It isn't necessary to do this as long as the MFT Reserved Zone is adjacent to the $MFT and contains sufficient free space for the $MFT to grow into in a contiguous manner. Since the MFT Reserved Zone is by default 12.5% of the drive, this leaves sufficient free space. For example, I have a 300GB drive. Intially formatting the drive results in a $MFT that is less than 1MB in size. Populating this 300GB drive with approximately 1 million files and 4.5 million fragments results in an $MFT that is 1GB in size - containing approximately 1,105,000 records (a single MFT record is a fixed size and can only hold a certain amount of information, so there can be multiple MFT records for a single file - which is something that is common with a heavily fragmented files). This would seem to be the type of scenario that you would want to pre-extend the $MFT to contain this large number of records. With all of these files and fragments and and the size of the $MFT growing from the initial 1MB size to the current 1GB size, guess what - no $MFT fragmentation - because of the MFT Reserved Zone and NTFS being able to allocate additional space for the $MFT by growing contiguously into the Reserved Zone.

From a pure technical perspective, this "feature" is absolutely un-necessary.

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[Mastertech]

No it only recommends pre extending the MFT reserve zone when the size of the MFT is anticipated to go beyond the reserve space.

[ghayes]

No it only recommends pre extending the MFT reserve zone when the size of the MFT is anticipated to go beyond the reserve space.

The MFT Reserved Zone is dynamic in nature. It can fluctuate every time that the drive is mounted. The only way to "pre-extend" the MFT Reserved Zone is via the use of a registry key. If this registry key is not present or the value is set to 1, then the Reserved Zone is set to be 12.5% of the drive. If the value of the key is set to 4, then the Reserved Zone is set to be 50% of the drive. The ONLY values accepted for this registry key are 1,2,3,4. Please note that resizing of the Reserved Zone (done on drive mount) is dynamic and the Reserved Zone can either increase or decrease in size - WITHOUT HAVING TO REFORMAT THE DRIVE.

Only if you artifically extend the $MFT by creating additional records do you have to reformat the drive in order to recover this space. They ARE artifically extending the $MFT. It is absolutely NOT necessary as long as your defragmenter bothers to ensure that the MFT Reserved Zone is cleared.

Maybe that is what this feature is really about. If they don't take the time to clear out the Reserved Zone when defragmenting, then NTFS is prevented extending the $MFT contiguously.

NTFS should really be allowed to to do what is inherent in it's design.

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[Mastertech]

Yes but Dynamic and Contiguous are two different things. If NTFS resizes the reserve space with disregard for it being contiguous, then pre extending the the MFT to be contiguous is valid.

[OxideNOS]

just compared diskeeper 10 and perfectdisk 7. Got good results with Perfectdisk 7, so uninstalled diskeeper 10. Try both and see for yourself.

[Wordsworth]

just compared diskeeper 10 and perfectdisk 7. Got good results with Perfectdisk 7, so uninstalled diskeeper 10. Try both and see for yourself.

I've tried O&O 8, PerfectDisk 7, and now Diskeeper Pro Premiere 10 and I've chosen Diskeeper Pro Pre. 10. Why? O&O 8's best defrag method (Complete/Access) takes too long, yet it's very thorough. PerfectDisk 7's Smart Placement strategy aims at preventing defragmentation, but I want a defrag strategy that focuses more on the quickest access or the fastest reading by the hard drive. PD does NOT lay out frequently *used* files but frequently *MODIFIED* files. There is a significant difference. I've actually had PerfectDisk 7 benchmark slower compared to O&O 8's complete/access method. Plus, I've always been a little suspicious after Raxco tested PerfectDisk 7 against the weakest defrag strategy in O&O 8--Stealth strategy. A more fair test would have been to test using O&O 8's Space strategy.

If I use O&O 8's complete/access method, even after several uses it's still too slow for my taste and that's why I've moved to Diskeeper (Premiere version). Once Diskeeper uses I-FAAST the following time is much quicker though I get similar performance results between O&O and Diskeeper. Diskeeper's I-FAAST strategy aims at faster read access too and avoids being fooled by certain conditions. I believe I-FAAST takes about one week to learn your computing style and adjusts the drive appropriately--I noticed immediate improvements. Their automatic scheduling is easier and comes with more choices. Bottom line--my system feels quicker and stays quicker with Diskeeper Pro Prem. 10. Benchmark test with Diskeeper shows improved results that stick around. Here's an indepth PDF paper on I-FAAST:

http://www.diskeepereurope.com/assets/_pdf...calibration.pdf

[srdiamond]

O&O 8's best defrag method (Complete/Access) takes too long, yet it's very thorough. PerfectDisk 7's Smart Placement strategy aims at preventing defragmentation, but I want a defrag strategy that focuses more on the quickest access or the fastest reading by the hard drive. PD does NOT lay out frequently *used* files but frequently *MODIFIED* files. There is a significant difference.

According to the O&O Defrag 8 help section, Complete/Access AND Complete/Modified are BOTH suitable for file servers. However, for personal workstations, O&O recommends Complete/Name.

"The COMPLETE/Access method is suitable for servers and workstations. Please bear in mind that this method uses a lot of central memory due to the reorganization of your system."

"The COMPLETE/Name method defragments your files and also reorganizes your file structure. Although this method is slower than the STEALTH and SPACE methods, it guarantees maximum performance for the read access to your system.

"The COMPLETE/Name method is particularly suitable for system volumes."

Stephen R. Diamond

[srdiamond]

Yes but Dynamic and Contiguous are two different things. If NTFS resizes the reserve space with disregard for it being contiguous, then pre extending the the MFT to be contiguous is valid.

If NTFS resizes the reserve space but disregards contiguity, isn't the solution for the defragmenter to insure contiguity, not to extend the MFT? Why should insuring contiguity require giving the MFT more space, if there'sufficient space already?

[Wordsworth]

According to the O&O Defrag 8 help section, Complete/Access AND Complete/Modified are BOTH suitable for file servers. However, for personal workstations, O&O recommends Complete/Name.

"The COMPLETE/Access method is suitable for servers and workstations. Please bear in mind that this method uses a lot of central memory due to the reorganization of your system."

"The COMPLETE/Name method defragments your files and also reorganizes your file structure. Although this method is slower than the STEALTH and SPACE methods, it guarantees maximum performance for the read access to your system.

"The COMPLETE/Name method is particularly suitable for system volumes."

Stephen R. Diamond

I've used each complete method found in O&O 8 --all require substantial time to complete even after first-time use. According to the computer's "feel" and benchmarking results, the Complete/Access method produces the best result on my system over and above complete/name and complete/modified. But with Diskeeper's I-FAAST technology, I get great results even better than O&O 8's complete/access method without the wait (after I-FAAST's first time use). According to the white paper previously posted, Diskeeper is the first defragger to combine the calibration results from the disk's physical architecture with the logical placement of files including free space. This method is a step above all smart placement or access strategies used by other defraggers such as PerfectDisk 7, O&O 8, Vopt XP 7.22

[Mastertech]

If NTFS resizes the reserve space but disregards contiguity, isn't the solution for the defragmenter to insure contiguity, not to extend the MFT? Why should insuring contiguity require giving the MFT more space, if there'sufficient space already?

Giving it more space while the space is still contiguous. It is a preemptive measure.

[Andy07]

great stuff!

[ghayes]

Yes but Dynamic and Contiguous are two different things. If NTFS resizes the reserve space with disregard for it being contiguous, then pre extending the the MFT to be contiguous is valid.

Since NTFS avoids putting files INSIDE of the Reserved Zone unless it gets into a low free space condition, the likelyhood of NTFS dynamically "shrinking" the Reserved Zone and NOT leaving enough free space to extend the $MFT contiguously is quite rare.

Again, the Reserved Zone and NTFS dynamically growing/shrinking the Reserved Zone is completely different than pre-creating records in the $MFT and artifically extending the $MFT.

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

Giving it more space while the space is still contiguous. It is a preemptive measure.

And quite un-necessary - especially as the majority of the $MFT (all but the first several clusters) can be defragmented online under Windows XP and Windows Server 2003.

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[srdiamond]

Giving it more space while the space is still contiguous. It is a preemptive measure.

Could you expand on this? What exactly is being pre-empted? Since you seem to agree that 12% of the disk space reserved for th MFT suffices, it isn't the condition of lack of sufficient space that is being pre-empted. So what is? Or is it that you actually disagree that 12% of the disk suffices to hold the MFT?

An additional point. The ability to expand the space reserved for the MFT is inherent in Windows and can be accomplished by a simple registry change. If this is important or even potentially valuable, it could easily have been implemented in an early version of Diskeeper. Then why did the developer wait until version 9 to included it? It has the smell of a measure tacked on to justify a thin version 9 release. All the changes in version 9 could easily have been included in version 10, which would have made the steep price for that version a little fairer. I have the sense that Executive Software is suffering financial problems, perhaps due to the sharp competition from Raxco, and was looking for an excuse to milk consumers for more cash.

MasterTech, from your web site I see that you are quite capable of seeing through the Firefox hype. Why don't you apply your critical abilities to Diskeeper? :|

[Mastertech]

Apparently Diskeeper isn't doing bad enough to send reps to the Neowin Forums to drum up support. There is no hype Diskeeper flat out works. I like to recommend solutions that require the least amount of user interaction as possible. While Perfect Disk will defragment your HD fine, the Diskeeper interface I have found is easier for end users and the automated features flat out work. I can come back in two years to a client and the drive is still defragmented. There is no hype. I don't tell people who are using Perfect Disk to switch ect... I simply recommend to new users what I prefer. I have thousands of clients I deal with and whatever saves me and them time and money is worth it IMO. Diskeeper has also been proven to be ridiculously reliable. I just don't have any problems with the product. Anyone who has used it knows what I am talking about. But if you are happy with Perfect Disk fine.

[Carlos11]

I'm a registered user of Diskeeper 9 Pro.

Very reliable indeed.

But the last version has truly one new feature: the I-FAAST technology.

However, Diskeeper 10 Premier Pro is insanely expensive.

I can purchase a new HD instead. Madness!

Finally, Diskeeper Corp. does not even provide a special price for current consumers. Just an upgrade to the Pro version which is basically the 9 version with a different "dress". C'mon! :(

[ghayes]

Apparently Diskeeper isn't doing bad enough to send reps to the Neowin Forums to drum up support.

Just in case this is a "dig" at me... I'm not a member of this forum to "drum up support". In my role as a Microsoft MVP for Windows File Systems, I participate in a very large number of online forums - including Microsoft's public newsgroups. I also also travel and speak quite a bit to various groups on Windows File System internals and performance. I attempt to educate people on Windows File Systems and clear up any myths and mis-perceptions that may be floating around :)

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[srdiamond]

Do you receive any money from Executive Software for promoting their product?

I currently use O & O, simply because I find it's screensaver mode particularly convenient. I'm not concerned with eking every last ounce of performance from my hard disk, since I'm not a gamer; my disk is half empty and is rather small, anyway; I'm sure the Windows defragmenter could handle it easily. So it isn't a matter of what works for me. Actually, PerfectDisk happens NOT to work well for me, because of its conflicts with StyleXP. This would be my major criticism of PerfectDisk, actually; it seems to have more conflicts with other software than competing products. But if my disk were hard to defragment or I was operating a server, PD would be my choice. I have a license for O&O Defrag 8, PD 7, and DK 9, so I'm not resolving any cognitive dissonance about my choices, except perhaps the choice not to spring for another $75 for the DK 10 upgrade.

I am, however, interested in the technical issues, and I consider a discussion of the virtues of different defragmenters are interesting entry point to various technical issues. The choice between them to me is a matter of convenience at the moment. But I was disappointed that you didn't respond to the substantive question I asked.

[Wordsworth]

Do you receive any money from Executive Software for promoting their product?

I currently use O & O, simply because I find it's screensaver mode particularly convenient. I'm not concerned with eking every last ounce of performance from my hard disk, since I'm not a gamer; my disk is half empty and is rather small, anyway; I'm sure the Windows defragmenter could handle it easily. So it isn't a matter of what works for me. Actually, PerfectDisk happens NOT to work well for me, because of its conflicts with StyleXP. This would be my major criticism of PerfectDisk, actually; it seems to have more conflicts with other software than competing products. But if my disk were hard to defragment or I was operating a server, PD would be my choice. I have a license for O&O Defrag 8, PD 7, and DK 9, so I'm not resolving any cognitive dissonance about my choices, except perhaps the choice not to spring for another $75 for the DK 10 upgrade.

I am, however, interested in the technical issues, and I consider a discussion of the virtues of different defragmenters are interesting entry point to various technical issues. The choice between them to me is a matter of convenience at the moment. But I was disappointed that you didn't respond to the substantive question I asked.

I'm with you on having conflicts between Style XP and PerfectDisk 7. I'm a moderate gamer with 4 partitions on my drive. I've isolated the OS and games from each other, as well as another partition for music storage. For me PerfectDisk 7 had been the most reliable of choices. It did a better job with boot defrag. But I love O&O's regular performance results in versions 6.5 and 8.0, though I think DK Pro Prem. 10 is a better product. What's funny is I plan to buy Vista when it comes out and then I wonder who will do a better job defragging that OS? I guess we'll be talking about that too. I don't receive any money from these products.

Edited January 5, 2006 by Wordsworth

[ghayes]

Actually, PerfectDisk happens NOT to work well for me, because of its conflicts with StyleXP.

Yeah, StyleXP is a source of "pain" for some people. Here's the deal. PerfectDisk requires exclusive access to the drive in order to ensure that boot time defrag occurs safely. We do NOT want the drive being modified by another process while we are in the midst of doing the boot time defrag. If we detect that another process or driver has opened the drive prior to us starting, then we will abort our boot time defrag. The issue usually is that the 3rd party driver is set to start early in the boot process when it actually doesn't need to. There are other applications besides StyleXP that can prevent PerfectDisk's boot time defrag from running. For example:

Symantec - PCAnywhere - gernuwa.sys driver. If you change the start of this driver from boot to System, PCAnywhere still works - PD Boot time defrag works. Why does this driver need to start Boot if PCAnywhere still works correctly with this driver set to start System?

C-Media - Audio driver - CMUDAX.SYS. Why the heck does an audio driver need to open the drive for write access?

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[srdiamond]

Greg,

Why is it important that another program avoid accessing the drive during a boot defrag, whereas such access during an ordinary defrag is harmless?

Also, does the fact that competing programs allow other programs access during their boot defrag procedures derive from their defragging less thoroughly or, in your opinion, are they just less concerned with safety than is Raxco?

[ghayes]

Greg,

Why is it important that another program avoid accessing the drive during a boot defrag, whereas such access during an ordinary defrag is harmless?

Also, does the fact that competing programs allow other programs access during their boot defrag procedures derive from their defragging less thoroughly or, in your opinion, are they just less concerned with safety than is Raxco?

It's not that we care if they access the drive for read. We just don't want them modifying the drive in any way. Defragmenting online is a little different than defragmenting at boot time - where minimal OS code is loaded. Our view is to take no changes. I can't really vouch for the safety of other boot time defragmenters as I don't have access to their code. Assumption would be that they are doing things in a safe manner. If they are okay with something modifying the drive during boot time defrag, then they are okay with it.

- Greg/Raxco Software

Microsoft MVP - Windows File Systems

Disclaimer: I work for Raxco Software, the maker of PerfectDisk - a commercial defrag utility, as a systems engineer in the support department.

[srdiamond]

Here's a problem I'm having with Diskeeper; I wonder if anyone here knows the solution, as I haven't been able to get a response from Executive Software's self-described excellent tech support.

I installed the DK 10 trial. To avoid confusing myself, I uninstalled DK 9 Pro, which I have a license for. After deciding not to buy the upgrade to version 10, I uninstalled it and tried to reinstall version 9, but I get the following message, even when turning all background applications off and all unneeeded services off:

Error 1402. Could not open key:

UNKNOWN\dFRGsNAPIN.dFRGsNAPIN.1\CLSID. Verify

that you have sufficient access to that key, or contact

your support personnel.

I have access to any key that might fit this description.

Does anyone know what this might signifiy? I tried reinstalling the version 10 trial to see if it would clear up any problem. Version 10 easily installed and uninstalled, but the problem with version 9 (which I am installing from a CD-Rom) persists.

[Andareed]

contact your support personnel

[srdiamond]

Who would that be, for a home computer, using a program advertised as appropriate for home users?