* So far, I can say the reverse of what you say. I tested that myself to see if my defrags take as long as yours, they do not... we did that at NTCompatible.com in fact. I remember doing it well, because I was like "He may have a point in it taking forever" & mine did not! I had already done a defrag by that point, & it was quick on the run I tested for you! It took me 8 minutes to do that one & I had done one previous days before if memory serves me correctly!
So apparently, the difference may be in usage patterns, which would meake sense. If patterns differ from day to day, or there is no distinguishable pattern, SD will constantly shift things around during every defrag. I notice this pattern on my disks. And on a sidk like mine, where most of the files are considered "frequently accessed" by SD, this will cause a lot of rearranging in the defrag run, thus taking a long time. Whereas, on your computer, you seem to have a good pattern set, and thus not a lot of rearranging is going on.
Am I correct in assuming this?
Really now? Windows XP & 98/ME are very different. this is true, but the software technology in the 98/ME defrag is alot like Symantec speedisk and it should be! It's Symantec code, one that practices bands of use & counts of use placements in BOTH 98/ME's defragger & also Windows XP Speedisk 5.0!
They both use filesystem data for tracking files, especially LastAccessDateTime stamps & counts of use of a set of files. this is exactly what those defraggers use from the MFT$ in NT based Os' using NTFS at least, to get their information on WHERE to place those files. I wonder how many diff. users use those machines those folks are sitting on that have these LONG DEFRAG PROBLEMS! Because if it's more than one user, yes. I can see that happening! Diff. patterns of use are being done on those rigs with more than one users using them!
I agree. This may be why the Win98/ME defragmenters took such a long time also...
And, like I said: How many users use those machines? Bet their whole families use them & that makes XP BootOptimization/OptimalLayout go "willy-nilly" as well most likely contantly doing it differently for users that use diff. datafiles & diff. programs from one another.
Could be. But the difference between SD and XPs optimizations is that XP only performs optimizations every three days after compiling its list for layout.ini, and only performs these optimizations at boot time and during idle times, so as not to affect the users time on the computer.
SD on the other hand, rearranges everything at the same time as defragmenting, which takes a long time. XP seems to take like a minute or so to do its stuff, and will quit when interupted until the next available idle time. If you quit SD during a defrag run, everything may be scattered to the wind, and will have to start all over again on the next run.
Windows 2000 Magazine February 2001 issue, where Norton Speedisk blew AWAY both Raxco PerfectDisk & Executive Software's Diskeeper that's where, & you know that! PLUS, again, files out on the outer tracks are read/wrote to faster & this is an industry established fact!
http://www.winnetmag.com/Articles/Index.cfm?ArticleID=16442&pg=4
Theres holes in this article when it comes to the new DK 7 and XP.
DK 7 were more contrained to the W2K MoveFile APIs. There were much more files that couldnt be defragmented online. This has been greatly alleviated by the XP APIs. That, plus DK 7 runs much better at clearing freespace and defragging all the files in one run than DK 6 did. Thus, IMO, the article isnt valid towards DK7 and Perfect Disk 2000 when it comes to being run on XP.
Now, heres what the article says about Speed Disk...
"In my tests, Speed Disk's consistently solid performance distinguished the product from its competitors. I started Speed Disk on the 8.4GB SQL Server database, and the product finished the job with impressive results,
although the defragmentation took twice as long as a Diskeeper run." (So again, I'm not the only to notice this
)
"One problem I noticed during testing: Speed Disk consistently broke a previously contiguous MFT into at least two fragments. On two occasions, the software divided the MFT into several fragments. I called Symantec to troubleshoot this problem, and an engineer told me that Speed Disk commonly breaks the MFT into a few fragments in the process of consolidating files on a partition. He also stated that Speed Disk's file-placement strategy always causes the MFT to fragment into two pieces. This explanation struck me as odd. However, Speed Disk did a good job at overall file defragmentation, so an additional MFT fragment didn't overly concern me."
"The obvious advantage to this approach is that it eliminates the need to reboot a system and perform an offline defragmentation pass. The disadvantage is that OS changes could potentially render Speed Disk unusable until Symantec revises the code to accommodate the OS changes. Apparently, Symantec thinks that the performance and usability enhancements that its programmatic approach provides are worth this risk. So far, the gamble seems to have been worthwhile. However, because of code changes in WTS, Speed Disk doesn't currently support that platform. At the time of testing, Speed Disk wouldn't run on a Win2K system running Terminal Services."
Now, the article also says some good things, mainly that SD defrags near capacity disks more thoroughly that DK 7. I will agree with this aspect. While DK 7 will defragment >10% freepace drives with little to none fragment, it may still have a tendency to not defrag the freespace. I believe it will just defragment the file in place, and leave it there (for instance, 100 -600 MB movie files or something to that effect). SD will do a better job in this area, but at the cost of taking a long time to defragment.
Now, my question is, where does it say that it "whales" all over the other products as far as making your disk faster? The article states that SD attains to optimize the disk for performance and slower future fragmentation. But it never states benchmarks, or even tells if theres a noticable disk performance speed boost after defragmentation over the other products.
Same idea of bands of use placements of files in their defragger, correct? That is also what the BootOptimization/OptimalLayout features does every 3 days but, they do differ in their ideas of placements on Pagefile.sys & MFT$ and their method? SLOWER, no question, but putting it into the middle of the drive!
Its not the same idea. It uses layout in conjuction with prefetching. When the app is opened, XP monitors the process, noting how much resources is needed when opening that app, etc., and records it in a .pf file. Then it lays out the files needed where it knows where they are, so that future app executes quicker. This is a very general definition, but it kind of shows that the process is different. I think MS is considering good organization over optimal disk placement on the disk. Its quite possible that XP is able to read the files better in the way that it organizes them. This is why they told the defragmenter companies to lay off those files. Which Norton chose to ignore.
Yes, the middle of the disk is slower, but only nominally. I've done a lot of research the past couple of days on this very subject, and every single trustworthy authority said yes, by definition, the outer disk is faster, but not appreciably. In other words, in most situations, you wont notice the difference.
If you modify a 350-engine to give it 5-10 more horsepower, its doubtful that you would notice the difference. But if you modify the carburator to get better gas mileage, your more likely to notice the benefits. I think this is the approach that MS is taking...
You say it's only a nominal increase. I will take that "nominal increase" & alot of those types of increases have had this very machine beat out a Dual CPU Palomino 1.2ghz, & a Dual Athlon 1.4ghz rig, pretty handily along with a Single Athlon 1.7ghz splitting tests right down the middle!
This machine is only at 1127mhz. Explain that! I will for you: By using those 'nominal increases', but ALOT of them, for better (far better) performance than without them!
Those machines I tested against, with 3 diff. benchmarks programs no less (SiSoft Sandra, DrHardware, WinTune all of currrent models) were raw, untuned ones, but FAR STRONGER hardwares Yet, I beat them & folks were like "WoW" ²!
Tunings those 'nominal increases', got me the win. Everytime, against 3 diff. machines expected to beat me! They accumulate to a sum that is greater than the whole of their parts apparently because I too, was VERY surprised!
My friend has a 1 ghz pentium. Mines an Athlon 700. I can beat his computer in almost any benchmark because I have "tuned" my machine more than his, and know the bottlenecks of mine much better than he knows his. This could be due to a number of things, like cruddy parts, cruddy OS, cruddy lack of optimizations.
When doing a comparison, you need to do it on your won machine with both products, and do it in an unbiased manner. Thats a hard thing to do, I know.
The point is, I *have* tried both DK 7 and SD (with your patch) on my system. I havent noticed any increase or decrease in speed between the two products, with the exception of defrag run times.
You have admitted that you havent run DK 7 yet. So how can you say whether its faster or slower?
More in next post....
