Keep The Window's Performance Runs OK

[aramico]

Dear Friends,

Nice meeting you all here..., Im a new comer here..... hope that we'll get better knoeledge n performance day by day here...

I have one main question, How can I use the backup in windows xp maximumly... ?

We know that we get the best performance at the first time after windows and any other needed aplication installed. Any model of virus, etc will drive our system's perfomarnce goin' down.

How Can I use the backup in order to bring my system "totally" same as the first time... ?

And how can I protect the backup mechanism without anti virus or something like that..., in the same time.. I do ussually patch the neccessary.

Regards,

Aramico

[Zxian]

If you're looking for an imaging solution, Acronis True Image is one of the best products out there for that.

If you want to go down this route, then I'd suggest creating a second partition for your personal files, and keep the system drive for system files and program settings.

[Jeremy]

I second that. But before you image your drive, since you want to improve your system's performance, you may want to take care of the following:

1. Startup entries

Start > Run > "msconfig" (without quotations) Check off anything you don't use on a constant basis. Make sure you know what's running in the background.

2. Services, they don't all need to be started and set to automatic.

3. Set your page file to 1.5 times the amount of RAM you have and have the initial and maximum value set the same as to reduce excessive fragmentation (since file thrashing will occur between your page file and RAM).

4. The size of your harddrive also bootlenecks your system performance overall. (Something to consider for the future)

5. Defragment your harddrive with a good defragmenter (Diskeeper 2007) Automatic, you don't need to manually defrag or schedule anything... ever...

6. Use a freeware program such as CCleaner to remove junk files/temp/cache/etc. About 9-10% of all computer users worldwide use this.

7. Give your registry a gentle cleaning with a safe program (CCleaner does a good job, also jv16 PowerTools.)

8. Be secure, malware/viruses can affect system performance as well. (Router, Kaspersky/NOD32)

11. Also, turn off all Visual Effects, if not all, the "heavier" ones.

[Zxian]

2. Services, they don't all need to be started and set to automatic.

5. Defragment your harddrive with a good defragmenter (Diskeeper 2007) Automatic, you don't need to manually defrag or schedule anything... ever...

11. Also, turn off all Visual Effects, if not all, the "heavier" ones.

Services - Do we want to go here again?

No, but seriously, if you're not willing to do some non-standard troubleshooting if something doesn't work (due to services), then don't disable them. The performance gain that you get is really negligible.

Defragging your drive won't help anything in terms of the image. The files will be stored in the image the same way regardless of their fragmentation on the original drive.

And I like my visual styles. I have never been bothered by my computer being "slower" because of it.

[Jeremy]

Either disabling some Services or simply setting them from Automatic to Manual, you will notice rebooting takes less time and on older PCs performance increase is noticable. For the typical home user, a little time to get to know your Services is actually something I recommend. There are lots of good guides out there for Services, and then there is trial-and-error, which isn't bad at all. A few reboots later and you've already accomplished something, and that takes what, 10 minutes?

Computer Browser, Server and Workstation, Network Location Awareness are needed for networking, but if you run only a single PC, then you can disable these and you will notice a difference after you reboot.

Windows Image Acquisition Service, when enabled, allows you to view your webcam within Windows Explorer. It is needed for webcams and digital cameras, but even when disabled, they function normally. I've tested it and am letting you know.

Defragmenting, yeah, you're certainly correct about that, but Diskeeper is still very healthy for your system to have it installed.

Visual Effects, ok performance wise they don't do much when turned off, but they use a lot fo animations which in the grand scheme of things take several seconds overall to display, I don't need to see a window minimize, when I click Minimize, I expect it to jump down there instantly.

[CoffeeFiend]

Zxian: strongly agree on all points. And diskeeper is pretty much the very last defragmenter I'd use...

[Jeremy]

And diskeeper is pretty much the very last defragmenter I'd use...

Have you tried it? Do you know that version 2007 is out now and it has eliminated the need for scheduling entirely? Do you know that it's automatic and runs in the background when it sees fragmentation and waits for idle resources before performing its tasks? You know the guys at Diskeeper wrote the original defrag API for Windows which is what all the other defrag programs are based off of? Why is any of this not good? Please let me know why this defragmenter is not good. I'm very curious.

Edited November 2, 2006 by Jeremy

[CoffeeFiend]

now and it has eliminated the need for scheduling entirely? Do you know that it's automatic and runs in the background when it sees fragmentation and waits for idle resources before performing its tasks? You know the guys at Diskeeper wrote the original defrag API for Windows which is what all the other defrag programs are based off of? Why is any of this not good? Please let me know why this defragmenter is not good. I'm very curious.

Couldn't care less if it needs scheduling or not. And personally I don't want something defragging in the background all the time, so I wouldn't use that (I'm very suspicious at how good it is at telling how "idle" a machine is - especially WRT I/O, not CPU-bound stuff).

As for the API. They didn't write an API for windows (just like Citrix didn't write the TS API). They wrote a defragger that hooked directly into the kernel (replaced it). And MS licensed some old version of that to make their own dumbed-down useless defragger from it. MS made their own API (on request of Executive Software) for NT4 to accomodate defraggers so they wouldn't have to keep doing that. Again, they didn't write any such API.

And being first means absolutely NOTHING about being good, much less the best.

It's about how good of a defragmenting job it does, and we've argued that over and over again (I'm sure there's plenty of talk about all this in that 13 page long thread). PerfectDisk and O&O are far better/more thorough (DK being pretty much dead last). I'm not going to get into that old "which is better" argument all over again. Again, it's been discussed for more than 10 pages already. It would be just as pointless as another "which is best IE/FF?" thread. People can look into it and decide what's best for them. I certainly wouldn't recommend DK though.

[Jeremy]

@Crahak

Yeah, 13 pages of people saying "<Insert name here> rules!" is not really informative.

I always like the try software before I go on a huge long rant claiming whether or not it sucks. Especially when most people are still probably using v10, when v11 is a lot different. Software should be given second chances.

Having said that, and us not being the people who coded the program anyway, who are we to say how good it really, least of all users who haven't used it (or newer versions)?

I've been using DK for a few months now, and if I have to choose to manually defrag my drives, sit around and wait for MB upon MB to be moved before I do use my PC again, or have intelligent software do all that in the background for me, without impacting my performance at all, then I think the choice is painfully obvious.

Downloaded movies open instantly, rather than seconds later. This is because DK defragged them before I decided to open. PerfectDisk or any other program wouldn't have done that and I would have had to schedule a defrag overnight after I went to bed.

You claim it sucks because you claim it isn't thorough. Well, have fun with other ones, then.

[CoffeeFiend]

@Crahak

Yeah, 13 pages of people saying "<Insert name here> rules!" is not really informative.

Perhaps 10 pages of it are that, but there's certainly some information there (there's even several posts by you that say more than that, so you should know). My point still stands (no point in duplicating that post here, that thread's already for that, and that whole discussion/argumentation thing already happened there... Opening a new topic for it/debate over that is useless)

I always like the try software before I go on a huge long rant claiming whether or not it sucks.

I have tried all versions of DK (including v11, build 608, server ed).

Software should be given second chances.

Hell, I gave Internet Explorer a second chance with version 7. If that's not getting a undeserved second chance, I don't know what is

Having said that, and us not being the people who coded the program anyway, who are we to say how good it really, least of all users who haven't used it (or newer versions)?

Mostly irrelevant. I didn't code most of the software I've used, and I can certainly tell some of it truly sucks. Actually, I'd say it's much the inverse. Those who've worked on it are most often too biased to start with to give balanced opinions (I'd like to see the CEO of DK say that another defragmenter is better, like there's a chance of that ever happening, even if it was true)

I've been using DK for a few months now, and if I have to choose to manually defrag my drives, sit around and wait for MB upon MB to be moved before I do use my PC again

Yeah, sit there and stare at it defragmenting. That's like loading up your dishwasher or laundry and sitting there while it washes and complaining you have to wait. Or perhaps you can schedule things. The average person working 8h/day 5days/week, plus around the same sleeping everyday, that means almost a hundred hours of time off that computer every week (not counting overtime, time out doing activities/shopping/etc). One would hope that's enough time to do the weekly defrag thing (not counting most can do background defragmenting too or use the screen saver mode). I certainly don't sit there and watch 'em defrag.

or have intelligent software do all that in the background for me, without impacting my performance at all, then I think the choice is painfully obvious.

It might be for you. Actually, it is for me too: it likely means higher ASIO latencies, intellisense lagging and such, slower DB queries (and inaccurate benchmarks), errors in DVB streams (very sensitive), and all kinds of such nice things, likely dropping some frames when capturing or even choppy HD playback and such. Yeah, the choice is painfully obvious alright.

Downloaded movies open instantly, rather than seconds later.

Actually, I've never seen a movie open seconds later on any machine. Fragmentation is mainly a non-issue here (it needs very few parts to start - so additional delay should only be a few ms at worst, and when playing mpeg4 around 1mbps it's irrelevant). It would take some VERY extreme fragmentation to make a difference of several seconds (if even possible). The main thing would be a matter of directshow filter configuration (what codecs handle what FourCC codes, priority flags, etc) and player choice if anything.

You claim it sucks because you claim it isn't thorough. Well, have fun with other ones, then.

Until it actually does a BETTER defragmentation job (DK v2007 still doesn't place MFT where it really belongs - and it truly does a big difference in speed, it still needs more free space to work, it still doesn't defrag all metadata, hibernate file and such), and has decent management options (PD Command Center rocks and is free) preferably at the same price points as the others and such (why not drop artificial limitations like the 768GB limit on the pro ed? If someone has any recent RAID array, they need the version that costs twice as much), I have no plans to spend a penny on it. I consider it to be a considerable downgrade from my existing defragmenter, why spend money on that?

[Jeremy]

Mostly irrelevant. I didn't code most of the software I've used, and I can certainly tell some of it truly sucks.

That's easy when comparing programs like cleanup utilities or audio editing software, sure. Defragmenters I've come to find are more difficult to compare to one-another.

Yeah, sit there and stare at it defragmenting.

I didn't say that I sit and watch. But not everyone has the same schedule. On top of that I want to point out something very important: PerfectDisk doesn't defrag on the fly (as we know), and many users do a lot is disk-intensive operations (video editing for example) and someone who spends 8 hours a day rendering will fragment their harddrive very quickly. With PerfectDisk they would have to leave the PC on overnight (that is if another person wasn't to use that PC for 8 hours throughout the night to do rendering as well) to defragment it. With Diskeeper, on a machine that likely has 1 or 2 GBs of RAM (if the PC is mainly used as video editing, it would need that much RAM), it would all be defragmented on-the-fly in the background, invisibly, and at no expense of the PC's performance. I have 2GBs of RAM as do many users out there and I can actually hear my harddrive writing data when dealing with several GBs at a time. I hear when Diskeeper defrags my files and I can do as much on my system as I want to and since the time I began using DK2007, I have not observed any decrease in my PCs performance at all, whatsoever.

It might be for you. Actually, it is for me too: it likely means higher ASIO latencies, intellisense lagging and such, slower DB queries (and inaccurate benchmarks), errors in DVB streams (very sensitive), and all kinds of such nice things, likely dropping some frames when capturing or even choppy HD playback and such. Yeah, the choice is painfully obvious alright.

Huh? What are you talking about? All of the above-mentioned are things that happen on an extremely fragmented machine, something which would never exist on a machine being defragmented by any software.

Actually, I've never seen a movie open seconds later on any machine. Fragmentation is mainly a non-issue here (it needs very few parts to start - so additional delay should only be a few ms at worst, and when playing mpeg4 around 1mbps it's irrelevant). It would take some VERY extreme fragmentation to make a difference of several seconds (if even possible). The main thing would be a matter of directshow filter configuration (what codecs handle what FourCC codes, priority flags, etc) and player choice if anything.

I don't mean to be insulting but I think you are just hearing yourself talk on that one. I've never known a movie (typical 700MB AVI) to open instantly on a PC unless previously opened or defragmented. I used to use PerfectDisk and most movies either downloaded or copied to the harddrive would be in hundreds, even thousands of fragments. I realize your first thought is "Drive failing", but I assure you this wasn't the case. Regardless, I will attempt to reenforce that claim by analyzing my files with PerfectDisk again.

Until it actually does a BETTER defragmentation job (DK v2007 still doesn't place MFT where it really belongs - and it truly does a big difference in speed, it still needs more free space to work, it still doesn't defrag all metadata, hibernate file and such), and has decent management options (PD Command Center rocks and is free) preferably at the same price points as the others and such (why not drop artificial limitations like the 768GB limit on the pro ed? If someone has any recent RAID array, they need the version that costs twice as much), I have no plans to spend a penny on it. I consider it to be a considerable downgrade from my existing defragmenter, why spend money on that?

And where does MFT 'truly' belong?

I've been using it for a while now and as I've been saying throughout this topic, I have noticed a difference in speed as DK2007 has been taking care of my movies and all other files. In the past, they've opened in a second or so. Now they open in 250ms or so.

As for free space, I have plenty of that, but nonetheless, a valid point. Perhaps things will change in the future in this respect.

[Jeremy]

I've just done a comparison between Diskeeper and PerfectDisk so here are the screenshots:

Diskeeper 1

Diskeeper 2

Diskeeper 3

PerfectDisk 1

PerfectDisk 2

PerfectDisk 3

The different graphical displays are fine, nothing wrong with those. I personally like PD's better, but I rarely look at the GUI for DK anyway.

The stats given to me by DK, some I find irrelevant, others are contradictory. For example, it says there are 15 fragmented files in total on my C: drive, and those fragments amount to 35, yet it says the average fragment(s) per file is 1. 15 goes into 35 2.333 times, so their math is worse than mine and I thought my math was the worst in the world.

But besides the stats, let's get down to the reason why defragmenters exist: To defragment our files. Now, how do we define "Thorough", really? Crahak? Zxian? Anyone? What is 'thoroughness'? In 8 versions of PD and 11 versions of DK, haven't they figured out how to "defragment" a file? Are there various extents for how well a file can be defragmented? If a file is in one piece, there is no problem, correct? But then doesn't any defragmenter do that? If thoroughness lies in where and by what manner the program places the file, then that's another story. But, IMHO, not a very important one.

[CoffeeFiend]

Defragmenters I've come to find are more difficult to compare to one-another.

Not really... DK never did as good of a defragmenting job. Simple enough.

I didn't say that I sit and watch.

sit around and wait

okie It should be trivial to manage to schedule some defragmenting job for anyone that reasonably computer literate (unless they have a very hectic work and sleep schedule, which is extremely uncommon)

And BTW, I do a LOT of video editing, and it doesn't fragment the HD quite as much as you claim. And that's a unusual scenario either ways (you picked the 0.001% of the population that might benefit from your no-schedule way, hardly representative of anything)

And you claim that your background defragging is magical and consumes no resources. That's very funny stuff to read over and over again. And that it uses some perfect algorithm for knowing when there are enough resources to do defragmenting. Somehow, defragmenting (which uses a fair amount of memory, and certainly uses a lot of the available I/O) doesn't affect performance AT ALL, WHATSOEVER as you claim. Aerial Pork, I say. And none of the things I mention are a problem, even on a "average fragmented" machine, but add if some background defragmenting kicks in at the same time, it's bound to happen (and again, there is NO such thing as a "perfect detection of idle resources" - especially when it comes to I/O). Sounds like you've never done those things before... Anyone who've ever tried that could tell you background processes, especially those that are I/O intensive like defragging WILL certainly do that. Especially things like DVB capturing. It uses not much CPU, and not that much I/O either, but just "sneeze by the PC" and your recording is all corrupted. Do I trust a defragmenter not to fire at that time?

I've never known a movie (typical 700MB AVI) to open instantly on a PC unless previously opened or defragmented.

I've never seen it take "several seconds" as you claim. Most of the time is taken by loading the proper filters and such things. Not by reading the headers (which is very very little of the avi file - check with filemon if you don't believe the obvious) I've never seen defragmentation make any difference in how long it takes for mpeg4 files to open ever, no matter which defragmenter was used.

And where does MFT 'truly' belong?

3GB further into the disk, to quote Microsoft. It makes an instant 5-10% performance improvement (again, their claim), which is not negligible (and it never "goes away" unlike making files defragmented). See their "NTFS Preinstallation and Windows XP" white paper if you don't believe me. This has been known for ages (that's why ppl use cvtarea when formatting), yet DK is the only major defragmenter (AFAIK) that doesn't bother to move it where it belongs.

Your next post doesn't make much sense.

Now, how do we define "Thorough", really?

I'll try. That it defragmented everything it should have (e.g. skipping hibernate files and such things like DK does is NOT thorough) Having all files as close to possible in one chunk is the main goal indeed, but it shouldn't skip all kinds of stuff either (talking about DK here again). And I believe placement (not talking about the MFT) does matter, placing specific files in the faster-loading part of the HD to speed up windows does matter, and is part of the defragmenting job IMO. As for the MFT placement, there is no excuse not to do it. It's simple enough to do, and gives a significant performance increase.

That being said. I let people try/test/decide what's best for them. You're more than entitled to prefer DK for any number of reasons (or anything else). And to quote yourself: "I've discussed this literally to death to the point where I just can't be bothered anymore." We could throw this back and forth 150 mores times and it would change absolutely nothing. And my point still is, this really belongs in that 13 page long thread, no point in duplicating it. All that stuff was already argued about, and going over all this again is an absolute waste of time.

[Jeremy]

The whole placement thing takes me back to a discussion I had with Zxian some time ago. It concluded a trade-off between workload vs performance. The pro being having all files placed together starting at the beginning of the disk, in sequential order VS the extra time it takes PerfectDisk's SmartPlacement to actually do so (extra workload).

I also admit that DK lacks in the "offline" / "boot-time" department. It also doesn't do any special placement of files, other than the I-FAAST 2.0. Another thing, however, I would like to address.

And where does MFT 'truly' belong?

3GB further into the disk, to quote Microsoft.

If PerfectDisk follows this 'rule', then my screenshot shows something contradictory. I've done dozens of boot-time defrags on dozens of machines with PD in the past.

This screenshot shows that PD has placed the MFT zone near the middle of the disk, 50.82561 GBs (51 GBs) through the disk. This is a long way from 3GBs. Actually, if that remains true, then the MFT was already in the proper spot before I did the boot-time defrag.

Who believes anything Microsoft says, anyway?

What about older harddrives that are 2 GBs or less?

Even now that my MFT has been moved, I don't notice any absolute 10-15% performance increase, either. Maybe Diskeeper disregarded that because those claims are not true, or that they were previously believed to be, and no longer are?

Edited November 5, 2006 by Jeremy