Why doesn't Vista tell me that disabling PageFile can be beneficial as well? ;p

Boooooo.

 

because it affects superfetch and readyboost
also vista utitlzes the virtual much more than xp

 

For that matter, why doesn't it tell you that disabling Indexing can be beneficial, and that reducing your Shadow Volume and System Restore space can be beneficial, and why does it HIDE it instead of incorporating a handy slider like in XP? IMO, Vista is barely usable before doing these things, but M$ doesn't give us any clues, do they.

 

i wish i know hoow to reduce the system restore in vista 40gb is just to much

 

Check out the Tips and Tweaks Guide.

 

Because vista is not as smart as you.

 

lol the way liveonecare words it sounds like it does more harm than good. >.>

 

Hmm. I actually didn't see this tweak in the guide, so maybe it's not there.

(Or maybe I'm going blind).

Anyways, here are instructions.

 

I wonder... Could it be, perhaps, because they know more about how operating systems work than you do?




They're right. Don't disable the pagefile. It is not "beneficial".
Yes, there are a lot of urban myths about how it's best to disable the pagefile when you have lots of RAM. Guess what. They're just that, myths.

So why on Earth should Microsoft give you incorrect information?

 

Pagefile is there just in case your system crashes and it can hopefully recover after a reboot. I have 8gb RAM and 4gb Readyboost but still keep the pagefile active just in case. Performance wise there's no significant differences. What's a millisecs sacrifice in speed compared to a system recovery.

Do you turn off the Pagefile in XP?

 

The page file has nothing to do with system recovery. The pagefile acts as an extension to RAM stored on the HDD. It does not mirror the data in RAM (as that would defeat the purpose of having RAM, since RAM is faster than HDD), but rather stores data that exceeds the limit of your physical RAM space. Further, the data that ends up in your page file is by design the data that is in memory but is accessed the least or has been accessed the longest time ago, so it wouldn't do much to help with what happened at a crash (since the data relevant to a crash would likely be resident in physical memory, not paged out to your page file).

Disabling your page file won't help much for perf either, as your system doesn't want to use it. It is slower than RAM, but if you run out of RAM, it is a damn sight better to have some data bumped to the page file than it is to have your computer throw a fit about your memory being full. In fact, some programs will stop working correctly or at all if you don't have pagefile space (it is rare, but trust me, they do exist).

Edit:
As an analogy, think of RAM like your closet and the HDD like your attic/basement. If you have a huge closet and you can fit all your stuff in to it, you don't need to store anything in your basement or attic. However, if you run out of space in your closet, having a basement or attic will be useful, even if it is slower to get stuff from than your closet, because it allows you to keep more stuff in storage in total.

Now, if its mid July so you have all your summer clothes in your closet, if your closet catches fire (analogus to a system crash), you will lose your summer clothes. Your winter clothes may be in the attic or basement, but you can never get the summer clothes back. (This is why pagefile doesn't really help with system recovery.)

 

Jalf and Sirmetman,

You are both swimming upstream. Don't you realize that these tweakers know so much more about hop operating systems work than the folks who wrote it? Don't you realize that these same tweakers have never had a crash since they disabled their pagefile, therefore it is perfectly safe to do so? Don't you realize that even though there have never been any real comparisons made to show how much this speeds up the machine, these same tweakers know it is much faster?

Gary

 

Wow, that is sarcasm, correct? I mean it has to be right?

I am still confused as to what theory these tweakers are even working with that would support a speed increase. I know page file is slower than RAM but it is much faster than not having it available at all. I mean having pagefile on just means stuff goes to disk vs being discarded. I don't know, someone please even explain the logic or lack of as to how this even in theory speeds things up?

 

ok guys i have 4 GB ram but 32 bit Vista...i have disabled virtual memory cuz i read that 4 GB systems dont need to use it and disabling it improves performance...can anyone plz confirm this

 

1st off you have not disabled VM as you can't. You have disabled pagefile part of VM. I will not over explain as I am no expert in computers. VM makes the application think the data it is accessing is contiguous even though it is not. It does this even within RAM not just disk (pagefile). The applications requires this to operate properly. No I don't think you should disable as having pagefile on only costs some diskspace. Turning off risks crashing your computer.

 

Being a laptop forum I'm surprised no one mentioned you need a pagefile at least equal (perhaps slightly more) to RAM if you want to 'hibernate' your computer.

Even if a machine doesn’t really need a pagefile (the processes use less RAM than there is in the system) you can still gain a small improvement in performance by swapping out unused processes.

Swap is useful as a safety net. If an application or applications start using up all your system memory (programs with memory leakage anyone?), you’ll notice that your machine will became slow. Without swap, the computer usually becomes completely unresponsive.

If you're using more than 500 MB of pagefile, your performance will drop significantly. I always recommend at least a minimum 'fixed' pagefile of that amount if you're not going to use hibernate.

Cheers

 

I used to recommend a fixed pagefile too.....

Until a month or two ago, when I actually did a bit of benchmarking on it.







Now I recommend letting Windows manage it. Guess what, it turns out that this offers better performance than when you set a fixed size.

Now can we please get this farce over with? Do not disable your pagefile, it is there for a reason. Do not set it to a fixed size, because Windows runs smoother when it is in charge.

 

Jalf I'm thinking now would be the wrong time and place for me to claim turning off pagefile allows 32bit Windows to use 4GB's of RAM?

 

You need a page file, especially if you work with programs or games that can use alot of ram. Bad things happen when you run out of ram and have no page file.

Also even if you have ram available there are a handful of programs out there that will use the page file anyways due to the nature of the program or the data. I assume this is where you get your "performance boost" because having no page file at all means it will use the ram instead. That or your going to crash, eather way tho not worth it in my opinion.

 

The only perfromance-related scenario where disabling your pagefilecan conceivably help is a temporary condition. If you defrag your harddrive, especially if your page file is fragmented, disabling the page file, rebooting, defragging, re-enabling the page file, then rebooting again may make a small difference in perf (above and beyond that gained by the defrag alone). However, I've seen plenty of claims to the opposite. And as noted, this is only a gain when defragging, NOT during normal operation. When using your computer normally, you should have your page file on unless you absolutely need every last MB of HDD. And if that is the case, you probably need to delete some stuff or pony up for a new drive.

 

1) let windows determine the size of Pagefile.
2) defrag regularly like everyday before shutting down.
3) delete your c:/windows/prefetch/*.* ==> shift/del regularly.

Your speed gains will be quite lovely and a faster bootup.

It is impossible for any OS to only use the Physical Memory at Level II where the SDRAM is now.

CPU cache is the fastest memory on any machine.
SDRAM is the direct relationship with the CPU and is second fastest.

The pagefile.sys on the harddrive is the sixth level of memory and the slowest.
But is very important.
The wiser ones above mentioned SWAP and it would be good to try to understand what that is.
An OS juggles needs and options while executing instructions.

Disable your Aero in Vista if performance is too slow.

As a Java programmer I find it quite simple to crash a program with no pagefile present. Install a few Adobe applications like Photoshop and Acrobat Pro and Dreamweaver and watch the page file become as active as physical memory.

It is humorous to listen to people with no system experience discuss the performance aspects of any OS. They prove within the first sentence they are just letting off steam and prove they are ignorant of what they speak.

If you want to know Windows Operating Systems better, install Linux on the same machine and get to understand how it works by using it.

You may never use Windows again. I have to at work, but not at home.

Vista is nothing more than a time out for Microsoft while they take their time getting Windows 7 ready.

Okay this is more than enough to rattle the cages of the ones that don't know an ArrayList from an Sort Method.

Blessings,
Chetanji

 

Yeah, it was VERY sarcastic. I thought it was obvious, but I probably should have been a bit more specific.

Gary

 

Are you certain you need it for hibernation???? I know you need a "hiberfil.sys" file that is roughly equivalent in size to your RAM size. But I hadn't heard of any dependency on the pagefile. I am not suggesting you are wrong, I have never disabled the pagefile, I think it's a stupid idea to do so.

Gary

 

We have all heard the CLAIM that it improves performance, but I have yet to see a single shred of REAL evidence of that. I've posed the question to proponents of this alleged tweak many times to show some proof. Not a single one has done so.

Gary

 

It is also a solution for global warming! [sarcasm!!!!!!!]

Gary

 

Good point! There are indeed times when you might want to disable the pagefile for a few minutes. Another such case is when shrinking the C: partition using Vista. Because the pagefile is marked as unmovable (as is another file who's name escapes me at the moment) it may prevent you from shrinking the partition. There is a thread here about that topic.

Gary

 

What, this isn't the Linux forum? I though we were talking about swap partitions. You're right I'm wrong, sorry guys. I agree, never disable the pagefile.

 

defrag the pagefile or the c drive. lol.
for #3, i thought you aren't supposed to delete Folder "ReadyBoot" and Layout.ini?

I knew this would be a heated topic again...as I have said before, I don't do anything extensive on my laptop. Light gaming such as Gunbound and Kartrider. Browsing the web, homework and school. I have 3gb ram and will never come close to using what? Hmm, 60%? I've never seen it go above 50% actually. That is why I have it disabled. Anyways, since you guys are saying its better for windows to manage it...I guess I will.

 

My head was about to explode, then you guys come in and set everyone straight. Awesome!

 

well i true the paper away but i did a experiment on the vitual memory on vista and how it would affect load times.
i dont have all time because it was not significant
the programs i ran where:
3dmark06
call of juaraz dx10 benckmark
vista start up time
blacksite area 51
i ran each program 3 times each to make sure there was no errors
my specs are in the sig. i have 2 hdd and i used and i have 4 gb of ram
main hdd 2nd hdd
1. 8196mb 8196mb
2. 4096mb 4096mb
3. none none
4. 8196mb none
5. none 8196mb
6. 4096mb none
7. none 4096mb

as i said. i threw the paper away and decided not to post it becausee the times was not significant i just ran these programs on lload times. and because the was barley anydiffence i got depressed and did not see if i got any performance in playing games or programs sorry guys

 

do what makes you happy. that probably has more of a perceived effect than either turning pagefile on or off.

 

Until Vista crashes taking with it some important piece of work all because it made you happy when you turned off the pagefile. Great advice. [shakes head]

Gary

 

The Placebo effect, yes its very real.

 

500mg glucose pills please!

 

coming right up, would you like a side order of Raid 0 with that? Guaranteed to make you think your computer is running 2x faster

 

There is no need to defrag the pagefile as it empties every time you reboot.

On a linux partition you do not need to defrag because the unix hard drives leaves spaces between files.
Windows NTFS hard drives need to be defragged regularly. This is common knowledge. I have learned not to laugh at people who do not understand basic facts. You should google "hard drive defrag" and look at some different ideas about the subject.
It will change your mind.

I know a few people that have accidently deleted these two and boot up is still faster than before.

The idea which you missed is to clean out prefetch every 10 to 20 hours of Web/Net use. It will improve performance.

This is simple, common knowledge.

May simple wisdom come to us all.
Blessings,
Chetanji

 

The unix file systems, you mean. And that doesn't entirely eliminate the need for defragging. Sometimes the gaps get filled up. Sometimes files get fragmented still.
It's usually not a big deal, and most people live with it rather than bothering to defrag.
But fragmentation still occurs.

Are you so sure?
NTFS also tries to leave space between files to reduce fragmentation.

Well, that's good.

 

Chetanji, the data within the pagefile won't get fragmented per se, but the pagefile itself can be fragmented. I've seen it broken in to 5+ pieces across the HDD, with other fragmented files mixed in between. For absolute best defrag, disabling page file may help during defrag, if your page file is non-contiguous. The benefit will be negligible though.

 

I have learned that too, so I promise I won't laugh when I ask what prefetch has to do with Web/Internet use?

Gary

 

o_o
I know that you are supposed to clear it out every so often to help performance, just not those 2 files.

 

I am going to get chewed over this but my view will not change as I have and am experiencing a benefit which is increased performance, I have 4Gb RAM on a 64Bit Vista Ult system and my pagefile is shut off. There is even a menu option which allows this which I have attached below.

To my defence, I lay forth the two following reasons:

1. I observed it visually. Many will say things such as the placebo effect and whatnot but we cannot forget that this is our first true indication of change. A great example of this is useless turbo memory and it was this that first allowed us to tackle the fact that it did nothing for normal systems with sufficient ram; and

2. In the very mechanics, it most definitely has to improve performance. First off, RAM is the fastest medium in comparison to a cache on a hard drive and SDRAM itself. Next we need to consider that disk activity of a pagefile HAS to slow your system down in comparison to the absolute use of RAM. This is much more obvious in a hard drive than a ssd but it is only simple logic when you think of a spinning hds process of retrieving info.

Now to discuss this process, I found that reducing it still allows pagefile to work and, to an extent, does take priority over ram. This is obvious if you monitor your ram usage with pagefile on or off.

Only turning it off will force you to use the ram you paid dearly for which is the speedier method.

It was only when I turned it off that I did observe my RAM usage to increase somewhat and my performance to increase as well. Can I measure this? I am not Trebuin and can only judge by what I experience. Trebuin, however has just proven an argument that many have had for a very long time when he shows his testing of 32/64bit as well as XP/Vista 32/64bit.

64bit is a quicker system, a fact which I knew from my own visual observations for several months now. Still, as with this pagefile argument, I had to endure several telling me if I cannot document it, it just couldn't be.

Point complete....the mechanics alone show that it has to result in a performance increase coupled with my visual observations show my confidence in my point.

Now, having said all that, do I think that someone shutting down pagefile without sufficient memory is probably being very careless? Absolutely !

I have gone to the extreme of what I do to ensure that I don't require pagefile for anything. I can run dozens of programs and video and sound files without a problem...thats pushing it I guess.

So we have point...counterpoint. I get to listen to a few say I don't know my @#$ from a hole in the ground now and, well, thats why I stay here...to learn and help as well as stand up for myself lol.

EDIT: In the end, I wonder if our panic of turning this off is simply an overabundance of fear and the thought of change, the "It can't be" theory if you will. Maybe a question to tackle by the opponents of my view would be exactly why they have a selection for 'No Paging File' if it is so dangerous not to have one in use. Similarly, I have heard tons of rederick and opinions as to how it can be so dangerous yet no valid posts that say anoybody has been affected or their system hurt as a result of turning it off.

We have tackled these same point/counterpoint threads for pretty much EVERY tweak in my Guide, a guide which I will concede has alot of testing and assistance by guys like Schud and others. Closing, this Tweak is there and the thread still gets great reviews worldwide.

 

the only problem i have with turning off the page file is if my computer gets plugged nut or i have to force shut it off i'll loss my info

 

When your operating system runs out of RAM, it crashes. Or your applications crash. That's the bad thing he's refering to. And often times they just crash to desktop with no error message.

It's not always the big programs that will crash. The small programs need RAM too

I personally have a pagefile for I do alot of testing with VMs.

If your computer loses power, you are gonna lose your data regardless of the pagefile. Programs recover your "lost" data from autosave files they periodically save.

 

I could see how an OS optimized to use an x64 proc (ie Windows 64 bit) would work faster than a 32 bit OS on the same proc. When optimized for 64 bit architecture, the OS can take advantage of that extra data width per operation to work a little faster.

That said, there are at least 2 (more) reasons why pagefile disabling is highly unlikely to help with perf. Firstly, pagefiles have been around for years, and in all that time, people still haven't been able to get hard data supporting it being faster. Secondly, the way the OS uses the pagefile makes it so there is no real logical reason why it would work faster.

Until I see a viable explenation as to why, from a technical perspective, disabling the pagefile would benefit perf, I'm inclined to disregard any claims without hard data supporting them.

 

Les,

You and I have discussed this before. I repeat my question "where is the proof"? Visual observation is not proof. Sorry. That is purely anecdotal. Where are the REAL NUMBERS? We have McGrady saying he sees no difference in performance, that's anecdotal as well. We did have one person say he did run the numbers, and saw such a minimal difference that he turned it back on and ditched the data. (To bad he didn't keep it.)

To point #2, you are talking in absolutes. No one denies that there will be some impact. But the point is that impact is immeasurable by a human hand and stopwatch. Therefore it, for all intents and purposes is 0.

There are two very good reasons that the "No pagefile" option is there. When you need to shrink the size of a partition with a Pagefile the physical position of a pagefile fragment may prevent you from doing so. Pagefile fragments are marked as unmovable. If you turn off the pagefile while you shrink the partition size, you get around the problem. Reason number two is in the case of multiple hard drives or multiple partitions. The option is there to allow you to specify no page file on one of the drives and creating a pagefile on another.

Regarding "yet no valid posts that say anybody has been affected or their system hurt as a result of turning it off." We HAVE had such reports. Just a few days ago we heard of a gamer who ran low on memory and had to shut down his game while in progress or risk having it die. How many of the other crashes and bluescreens folks talk about on this forum are the result of running out of memory? The point is there is no way to tell. Vista doesn't warn you with a message like "Hey idiot I am out of memory and can't page out anything to free up some since you disabled my pagefile." Nope it just crashes the app, no message no nothing. The gamers code was proactive in asking the OS how much memory was available. 99.999% of the apps out there don't bother to do that. Having a pagefile obviates the need to do so. Most programmers assume the request for memory will be granted. Why? Because they assume the OS has a pagefile so it can grant the request.

I am afraid you are going to have to endure one more... [big ol grin] It's true, if you can't document it, it doesn't exist. It's like any other legal case, it is up to you to document the case. I can't prove a negative.

I appreciate the props. Yes this thread deservedly gets rave reviews. I have taken exception with a couple of the tweaks and you have bent over backward to include my warnings or replacements when I have done so. This is the only tweak I really take exception with. The very fact that the "performance improvement" is undocumented and there is no audit trail to point back to this tweak when a crash does occur, in my mind make it useless at best and harmful at worst.

If someone wants to take the time and do some scientific testing on the tweak and can show a performance gain that is substantive enough to offset the potential for unreported errors, just simple crashes, I might change my tune.

Gary

 

This appears to be a jobh for the great Trebuin...

 

Les, let's think scientifically here. Can you or anyone else form a hypothesis as to why no pagefile would be faster? Complete the sentance, "a computer with no pagefile will run faster because...". Your analogy to 64 bit OSes isn't valid because you can easily complete the sentance in that case. "A computer with a 64 bit optimized OS on 64 bit hardware will perform better than a 32 bit OS on the same hardware because technical optimizations; for instance, doing calculations on long integers (integers over 32 bits in length), can be run in fewer operations." There is no analogous statement I can think of for the pagefile question.

 

Oh the scientific case CAN be made. But the question is how significant is that case. Here is the case: without a pagefile you will have less disk I/O. My premise is however, that since most of that I/O is carried on in the background and there is such a small amount of I/O, that it's effect can't be measured by a human hand and stopwatch. And if it can't, then it is effectively zero.

Gary

 

Agreed! I wish I had the time to put this story to rest. I am too swampped at the moment to take it on.

Gary

 

But with a pagefile, if your runtime data can fit in RAM, it will reside there for the vast majority of programs. Therefore, the statement doesn't quite pan out. With a pagefile, you don't see HDD I/O unless you are running short on RAM. Without a pagefile, when you run short on RAM, programs will just crash. Therefor, the only difference under this case seems to be that with a pagefile, your programs will be more stable, not crashing when you run out of RAM, while without a pagefile, if you fill up RAM, programs die. I don't see how that supports the idea that turning off the pagefile helps.