Wednesday, June 13, 2007

The Great Defrag Shootout XVI: O&O Defrag V10 Professional

It's amazing what a few changes can make. Version 8.6 of O&O Defrag was a bit frustrating because it refused to defragment files, and one part of the program didn't seem to know what the other part was doing. Version 10 fixes a lot of these problems. The main thing that's new is the snazzy new interface with a ribbon bar. More important, though, is the engineering under the hood.
O&O Software claim to be able to do a defrag down to 5% free space, and the on-screen prompts explain for more clearly what is going on and how each defrag mode works. This in itself has changed my rating from a "Thumbs Down" to a "Thumbs up". But the sense of control is still a bit "rubbery": you click on the stop button and it doesn't always respond as fast as you want it to. The system doesn't always tell you what it is busy with and what it's up to. If you want immediate defrags no matter how little fragmentation is on the drive, this program will frustrate you; but if you want a program you can install and ignore, this is probably the best.
A big weakness is that you can't use the "AUTO" defrag and the screen saver defrag at the same time. Every time the screen saver kicks in it stops the auto defrag, and I couldn't figure out how to start it again. So you have to choose between Auto Defrag mode and Screen Saver mode, but neither the screen saver module nor the main module provides any warning of this.
One feature that works nicely is the ability of the program to resume a defrag where it left off, which saves time and makes it more efficient. Unfortunately version 10 still has some raw edges. The Help button didn't work when I installed the program on drive D:, so I had to go looking for it to find out what is new in version 10. The help file lists 4 items, but not under "What's New" but in the next section.
The boot up defrag takes a bit too long on my machine, but I can't complain that my files are in a mess: they're not. The progam allows for various kinds of defragmentation plans, but puts all the free space at the end of the drive, so new files are slower. Many other programs (exclusing JkDefrag and Ultimate Defrag) do this too, so it's not alone in this respect. This program allows you to see where the files are, a feature it shares with Vopt. But it has boot time defrag and a screen saver option, something Vopt doesn't have. I'd rather spend $44.95 for this program than $99 for Diskeeper Pro Premier or $49.95 for Diskeeper Pro, but it's more expensive than Vopt $40 and PerfectDisk $39.99.

The Great Defrag Shootout: Part I | II | III | IV | V | VI | VII | VIII | IX | X | XI | XII | XIII | XIV | XV | XVI | XVII | XVIII | XIX | XX | XXI | XXII | XXIII | XXIV | XXV | XXVI | XXVII | XXVIII | XXIX| winner | all | why

21 comments:

Anonymous said...

So what's your final verdict on O&O Defrag V10 Professional?

Donn Edwards said...

"If you want a program you can install and ignore, this is probably the best."

Anonymous said...

I've been doing a bit of testing myself just today, after restoring my system drive from total loss caused by PerfectDisk's incompatibility with BestCrypt whole volume encryption. As it turns out, PerfectDisk (both v7 and v8) seem to ignore their file exception list, and they cheerfully defragged (thus destroying) an ultra-critical system file needed by the BestCrypt encryption software to mount the system volume at boot time.

I just finished testing O&O v10, and it has no problem leaving this file untouched once the file is placed on the exception list. For the record, Diskeeper also left the file untouched once it was on Diskeeper's exclusion list. Paragon's Total Defrag 2007 didn't seem to have any method to exclude specific files, so it shredded my test disk just as PerfectDisk had done. But again, the thing that PO'ed me about PerfectDisk was that it ignored its exception list, confirmed by PerfectDisk's own before/after fragmentation analysis.

So far, O&O Defrag v10 has both of my thumbs pointing in a generally upward direction. It seems to be much improved over the earlier versions that had driven me toward PerfectDisk (before I was using encryption software).

Anonymous said...

I'm new to your blog, good reading. How about every so often post a verdict stat sheet comparing all your test side by side? Have you reviewed IO Bit yet?

Donn Edwards said...

See Bestcrypt page
http://www.jetico.com/bcve.htm
point 19 under fixes:
"Solved conflict with Raxco PerfectDisk defragmenting software. Windows Vista Disk Management utility now correctly displays information about encrypted volumes. (13-April-2007 release, User Interface v.1.97.2, Driver v.1.49)"
No defrag programs affect the contents of a file, only its location, so I fail to see how any defrag program could affect the encryption. I use the exclusion feature and it honours all the files I have excluded.

Anonymous said...

Donn... I also saw the "fix" report on the Jetico page, and I have the latest version, but PerfectDisk nonetheless rendered my system disk unbootable. As for your theory about defragging being irrelevant to encryption schemes because the content of any file remains unchanged, I believe the problem comes from the fact that one specific file is simply moved. If I understand whole volume encryption correctly, the volume can't be decrypted until BestCrypt finds those sectors, but PerfectDisk moved them from their mapped location.

In my testing, both Diskeeper and O&O also shredded the disk unless the "bcldr.bin" file was placed on the exclusion list. PerfectDisk rendered the disk unbootable no matter what. According to PerfectDisk's own fragmentation report, this "excluded" file had six fragments before defragging, but was contiguous after the operation. I then rechecked the exclusion list, and it was still there. My next reboot failed. I can offer no further explanation, but I will simply add that any defragging pass may not necessarily move any given file, so this is a risk that may not always be realized.

By the way, PGP disk encryption also had the very same problem with PerfectDisk according to older postings in their forum, and they changed their software so it now includes a rather bulletproof "protector" service to prevent any programs from moving their critical file. Unfortunately, that same PGP service gets in the way of some backup software, generating "unreadable sector" errors during backups.

There may not be huge numbers of people who use whole disk encryption, so maybe this isn't a hot topic, but I just lost my system yesterday and did a fair amount of testing to reduce the chances that it will ever happen again. And completely unrelated to that, I happened to get a newsletter tonight with a link to your fine article, and I thought maybe I could shed some light on the one aspect of defragging that you neglected to mention ;-)

Donn Edwards said...

My testing does include encrypted files, but my tests were one with TrueCrypt, which creates a 4GB virtual drive "M:" on my machine. The only program to show any sign of a problem with it was mstDefrag 2.0 beta, and even then it was a display problem, and didn't trash any files.

I haven't used PGP Whole Disk encryption, but used PGPDisk a while ago, and haven't experienced any problems.

I still think the problem is caused by the method used by BestCrypt to create the encrypted files, and it strikes me that any system that requires one of its critical files to remain in a static location or fragmented is just being irresponsible, knowing full well that WDD or any other disk defragmenter could irretrievably screw up a user's data. I recall the copy protection system for Ashton-Tate's dBase III did something equally stupid several decades ago.

I'm not arguing in defence of PD, but I really don't see how you can blame PD for the bad design of BestCrypt. I use PD's file exclusion facility on two key files on my drive, and it doesn't touch them. I have no idea why PD affected your files, and would be interested to hear what Raxco support had to say. (Did you contact them?) PD has defragmented my M: drive on many occasions, as has UltimateDefrag, O&O Defrag, Vopt and most of the other utilities I have tested.

I'm glad that O&O Defrag has met your requirements. Please understand that the tests I have run have been on my own personal machine, and in some cases machines I work on. I do not have a whole bank of computers in a lab or anything like that.

Anonymous said...

A recent update of PD8 from Raxco mentioned the fixes addressed and included this statement:
• “PerfectDisk Client
1. Code was added to exclude certain SafeBoot encryption files. Moving these files could cause your computer to become non-bootable."

Too bad they didn't indicate if the update assessed any damage and fixed it. Seems like there is more than one issue.

Anonymous said...

Drive encryption software may modify the (Master Boot Record) to point to a file that is expected to reside at a specific location on the drive. Examples of encryption that use this mechanism are SafeBoot, PGP and PC Guardian.

A disk defragmenter may "move" this file during the process of defragmenting the drive. As a result of this file being "moved", the encryption software no longer finds this file at the expected location on the drive and will not allow the drive to be accessed.

Drive encryption software vendors usually recommend that a disk defragmenter be configured to exclude one or more specific files so that they are not "moved" by the defragmenter.

PerfectDisk 8 - beginning with Build 64 specifically excludes the Safeboot related files so PD will not "move" then online OR at boot time. For PGP, we have a registry import file that can be used. Raxco has also been in contact with PGP and have provided them with the necessary information for PGP to tell PD8 which files to exclude (in case their new mechanism of preventing access to this file doesn't work).

This is the first that we've heard of an issue with BestCrypt. Assuming that the BestCrypt file is excluded in PD8, PD should not attempt to "move" this file during an online defrag pass. The excluded file should show up as Access Denied. However, PD may attempt to "move" this file during a boot time defrag.

This is the first that we've heard of PD NOT honoring files on its online exclusion list. We'll look into seeing if we can re-create.

While we are researching BestCrypt, in the intermin we have created a registry import that will tell PD exclude bcldr.bin

http://ftp.raxco.com/pub/download/pd80/tools/BestCryptExclude.zip

This registry import overrides any files specified in the PD8 GUI exclusion list. You'll notice that with this registry key set, you can NOT unselect this file from the PD8 GUI exclusion list.

Encryption products such as SafeGuard Easy do things a little differently. To ensure that their critical file will not be "moved" by disk tools, they mark the clusters occupied by their file as bad - so that nothing can "move" the file from its current location (Microsoft's defrag APIs don't let defragmenters touch clusters on the drive that have been marked as bad).

This does not effect the encryption perfomed natively by Windows - which doesn't modify the Master Boot Record.

- 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.

Anonymous said...

Fascinating info! Amazing that I found this by way of a RAXCO blog! Gives me hope that they may actually get around to improving their product.

My main system is a Dell P4 Precision Workstation with SATA internal drives and a large external drive (USB connection) mixed NTFS and FAT32, multiple partitions on each. PD does a pretty good job (fast, complete) with these. I have experienced PD's problem with large files being split across Meta files on this system and PD’s inability to defrag certain Meta files. I also note the general unresponsiveness of their support organization. I have observed, however, that after a number of additional defrag passes (10, 12?), these files seem to get straightened out (defragmented) -- or at least not reported as fragmented.

I also have an old P3 system with Win98 and XP in dual boot mode. PD (installed under XP of course) choked badly on this system (and I removed it) but then (because all the partitions are FAT32) I ran (under Win98) an old Norton Speed Disk (SD) from 2002 (nothing newer, of course) and find this defrag process faster, more complete, and with nicer interface. For example, you can click on any sector and SD will tell you what is there. Make no mistake, I am NO FAN of Symantec, which takes over good software companies and then ruins their products (think Partition Magic here) but if you do not have NTFS to fuss with, an old version of SD might be worth a try especially if you need to work with Win98.

I think PD's strength is how it approaches Meta data even if is still a work in progress so I am keeping it on my main system.

Thanks for sharing your findings with us.

Anonymous said...

I really appreciate your detailed revies :)

But I find O&O a little confusing because of all the different defrag strategies. I'm afraid choosing the wrong one will have a bad impact on performace.

So basically, what is ideal to use on a my Windows system partition (C:) ?

The manual seem to recommend COMPLETE/Name, but I've seen others recommend Access instead. Or should I rather stick to the faster SPACE algorithm since it's faster, and can therefor be run more often.

(Im' confused)

Donn Edwards said...

My guess is to use COMPLETE/Name for all manual defrags, and SPACE or one of the other less thorough algorithms for automatic defragmenting.

I can't recall seeing any dramatic performance difference based on file position alone, because there is only one major seek time involved, whereas fragmented files have an additional seek time for each fragment.

Obviously a file in the faster part of the disk will load quicker than a file in the slower part of the disk, but I'm not sure whether the difference is that noticeable, given that it would be measured in milliseconds, not seconds.

Donn Edwards said...

GaryM wrote about BestCrypt. I have just received an email from one of the Raxco software engineers that sheds some more light on the matter:
>>>
An update for you regarding BestCrypt.

Even if PD has been configured to exclude the bcldr.bin file, online PD will still continue to defragment the file.

In order for PD to exclude a file, PD must get the fileid assigned to the file - which is done by "opening" the file by file name and retrieving the fileid - which PD then adds to it's excluded file list. BestCrypt has a driver that prevents PD from opening the file by name (and since other defragmenters defragment by file name - prevents other defragmenters from defragmenting the file online). Since PD can't open the file by name, the fileid isn't returned, there is no fileid to add to our excluded file list and PD will defragment the file (because PD is able to defragment without using filename). According to the folks at Jetico that I have been working with, Jetico "has designed BCVE so that it should not be sensitive to relocation of its bcldr.bin file that may happen after defragmenting software activity". However, we are trying to determine if what they say is true. Using the reg import for BestCrypt will prevent PD boot time defrag run moving that file (their driver doesn't protect this file at the point in the boot operation where CHKDSK/PDBoot runs.

I'm also trying to get them to use the mechanism built into Windows to prevent a file from being touched by a defragmenter. A simple solution from their end would be to use the MarkHandle feature available on WinXP, Win2K3, Vista and Server 2008. This is provided as part of the FSCTRL APIs and was put there specifically to mark files as "unmovable" so that they are not capable of being defragmented using FSCTL_MOVE_FILE (which all defragmenters should be using - PerfectDisk does). With MarkHandle set for a file, FSCTL_MOVE_FILE will not allow ANY defragmenter to "move" the file. PerfectDisk's boot time defrag also respects MarkHandle and will not "move" any file marked as such.
<<<

Donn Edwards said...

A further update from Greg Hayes at Raxco:

>>>
I've been working with the folks at Jetico (You can find me on their Our Gratitude page - I've been VERY impressed with their response and willingness to do what needs to be done to ensure safe interaction between their software and disk defragmenters) and they have released build 1.98.4 which contains a fix that will BestCrypt to work correctly regardless if any disk defragmenter "moves" the file bcldr.bin.

With the release of this build, you no longer need to configure PerfectDisk (or any other defragmenter) to specifically exclude the bcldr.bin file.
<<<

Anonymous said...

Hi,

I just installed v10 on my system, I used to have v8.6 installed. Actually I don't find much change between the two versions. And that's a problem with many commercial defragmenters : publishers feel the need for issuing a new version at least each year, but changes are usually barely cosmetic and only deal with the GUI, the settings and some advanced functions no-one ever uses. I have the same feeling about PerfectDisk. I stick with V6...

Anonymous said...

Donn,

Does your verdict "If you want a program you can install and ignore, this is probably the best." still fall true, even with PerfectDisk 2008 released?

Thanks!

Donn Edwards said...

This is a brilliant question, and I'm not sure if I have a good answer, but it's certainly worth considering.

I haven't used O&O Defrag since writing the review, so I'm going to have to try it out once more.

My guess is that it will amount to personal preference. Stand by ...

Unknown said...

Dear Donn,

Did you happen to test out O&O yet in terms of it's "set it and forget it" feature?

Thanks!

Anonymous said...

Hi,

I found that defragging on vista home premium with either the name/ access/ modified methods wiped out my system restore. Apparently if the system restore file are moved then the OS is unable to track them or something.
Can you replicate this problem or is it something specific to my config?
Have you noticed this problem with other defraggers?

Bye and keep up the good work :-)

TM

Donn Edwards said...

PerfectDisk has a different problem with the Vista system restore files and can't move them. Maybe its related.

Try turning off system restore and then do the defrag. You can always turn it back on again afterwards.

What does O&O Tech support have to say about the problem?

Unknown said...

i was majorly disappointed from o&o v10. the stealth/name-method does not work reliably. neither does it sort files correctly, nor does it sort whole directories. in my case it was supposed to sort db*-

files. it started with db8 and db9. then it inserted a whole bunch of root files until it resumed with db0 to db7.

other shortcomings of this program are the fact that the single file and folder defrag does not work when the gui is open. you have to close it manually and then choose the defrag option through a right-

click on the files in explorer. i couldn't find an option that allowed me to defrag files within the program directly. this is unacceptable for such an expensive program. also, you don't see a report after

such a defrag, so noobs will not even realize that nothing has happened if they run the single file defrag with opened gui. also, in that mode you don't have any control over the method. let's say you just

want to defrag your 20gb vista folder, then you don't get to choose whether the sorting is done alphabetically, by access, or by name, let alone if layout.ini is being used.

i have to say that the hdd i tried it with hadn't been defragged since i bought it two years ago. however, it wasn't in bad shape. it contained 200,000 files and folders, but had only 2000 fragmented files

which caused a fragmentation of 22%. i tried the space method at first, but that did little but to defragment 1500 of the smallest files, while fragmenting the biggest files even more than before. i don't

really care, whether a 100kb file has to be read in 10 seeks or 1. it's those gb-sized files with thousands of fragments that i care about and o&o's space method was not capable of dealing with them. after

wasting my time with that, i switched to the stealth/name method, just to find out that it ripped the folders apart by inserting totally unrelated files from other folders.

you already mentioned it in your article: thanks to the sorting, the free space is always at the inner circle of the disc, so newer files will always have slower access speeds. none of the defrag methods o&o

supports, has actually options to place the files on the opposite end of the disc.

i also realized that the program works highly inefficient. when i used the stealth/name method, some files in the windows folder that were placed quite early on the disc were followed by gigabytes of

freespace. instead of just moving that data temporarily to the end of the partition like an efficient defragger would do, o&o just moved this data a sector forward, writing new data on the now free sector

and then begin again. it was pushing these few hundred megabytes over several dozen gb of free space. no need to say that it took ages...

also, this put twice the wear and tear on my drive as needed, since every sector had to be accessed 4 times this way instead of just two times. oh yeah, after having to stop after 50%, i wanted to resume,

but the program started over...it did jump a few portions that it had defragged earlier, but not enough, since it apparently consolidated every little byte of free space that was created after a reboot,

hence killing its own achievement and basically starting from scratch.

all in all i have to say that o&o is nothing more than a shiny app, that is far too bloated (compare the installed size of 33MB to those of the many smaller tools of only 6MB, which usually also have more

features) with too few options.

i'm also with merleone. o&o has nothing new to offer. last time i used it, it was version 6. except for the single file defrag it has nothing new to offer. even if the engine has changed by now, i doubt it was for the better seeing all the problems i ran into...

anyway, thanks for your great tests. thanks to those, i found the ultimate defragger for my needs :)