Just wanted to say, I've been running with no page file on my system (see sig) for a few weeks now, and everything has worked just fine. I don't think it speeds up the computer any, but it seems to access the hard drive less, which is good because the hard drive noise kinda rattles my nerves. Not saying people should go buy a lot of ram and turn it off, just saying nothing bad happens if you do.
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
With Windows Vista x64 and 6GB of RAM do you need a page file?
- Thread starter Archaea
- Start date
DeaconFrost
[H]F Junkie
- Joined
- Sep 6, 2007
- Messages
- 11,582
That's not true, and shouldn't be stated as a fact. Just because you've avoided a crash so far, on one particular computer, doesn't mean this will be the case for everyone. Think of the myriad of configurations and software combinations out there. The bottom line is, it does nothing for you or your computer, so there is absolutely no point in disabling it. If the hard drive noise is bothering you, mount the drives better.
People need to get out of this feeling that just because they've changed a system setting from the default, doesn't make them special, doesn't give them that warm and fuzzy feeling inside. Just because an option can be changed, doesn't mean there is a reason to do so. I understand everyone wants to feel like a superuber computer tweaker, but what's the point when it yields no gain? Would you put a lot of time into modifying a cars engine, if in the end, it had the same horsepower, running at the same mpg?
TechieSooner
Supreme [H]ardness
- Joined
- Nov 7, 2007
- Messages
- 7,601
QFT
Well, what do you know? You only tested it.
There's no point arguing with some of these people with skulls 8" thick. They want to cling on to strange notions. It will work, and it doesn't crash the computer when it runs out of memory. There's little reason to disable the PF, but it also worked fine in *everything* I ran when I tested it too.
Well, what do you know? You only tested it.
There's no point arguing with some of these people with skulls 8" thick. They want to cling on to strange notions. It will work, and it doesn't crash the computer when it runs out of memory. There's little reason to disable the PF, but it also worked fine in *everything* I ran when I tested it too.
Well, since the tool I work on runs on 450 million computers, perhaps my perception of what happens when a user turns off the page file is slightly different. If it works for your configuration, rock on, I'm happy for you.
I wouldn't turn it off.
But I see a lot of Out Of Memory errors, that when I debug I have it check for if the pagefile is turned off. You might be surpised to note most of the time, the pagefile is turned off, and low and behold I ran out of memory...
This has been debated hundreds of times, and all it does is end in a stalemate. People will advise things that work for their workload, that don't work well with other people's workload. Search posts by me on pagefiles and you will find some horror stories.
Do what you want with your machine, and rock on. But don't be suprised if I advise against it. But I've also been told my skull is kind of thick.
This posting is provided "AS IS" with no warranties, and confers no rights.
TheRapture
Supreme [H]ardness
- Joined
- May 31, 2000
- Messages
- 6,885
Unless you are seriously low on disk space.....why turn the PF off? It does not hurt AT ALL to have it on. A power user should have tons of disk space anyway....right? If you can't afford to have a 1 or 2 gig page file, you need to upgrade a bit.
Some situations WILL have issues if it is off...however, almost NONE will have problems with it ON.
Leave it on.
Regardless of what most people think, the engineers at MS typically know alot more than the average forum user
Some situations WILL have issues if it is off...however, almost NONE will have problems with it ON.
Leave it on.
Regardless of what most people think, the engineers at MS typically know alot more than the average forum user
Well this is a large load of crap. I have used XP and 32-bit vista on the very same hardware (4GB) and vista is noticeably more fluid. I use xp all day at work and then come home to my lesser powered laptop and it's far more fluid than my xp machine. Vista is better at managing memory than xp is...stop humping xp and wake up.
Don't get me wrong. I've repeated it several times in this thread. I don't think there's a good reason in general to turn off the PF. My week-long session running *everything* I normally and not so normally run had no problems with Vista x64, 6GB and no PF. It does work pretty well with enough memory in the system. *Everyone* probably can't, but *some* probably could run without a PF and never have problems depending on what they use.I wouldn't turn it off.
I can only think of one case when I would want to disable the PF: on a mini notebook with a SSD, or running an OS from a USB stick.
Interesting thread, i've seen a few of these over the years and it's interesting to see that little has changed with memory management.
I'd say if you're going to state anything as fact like dummy pagefiles are being created in certain circumstance, to list those circumstances and all the requirements we need to test and confirm you're right.
I'd also say that testing something for only a week with only average use is unlikely to show any problems, it's always going to be an extreme circumstance that knackers you if such a thing occurs, and your testing simply isn't covering that, maybe applications have memory leaks and would reach high memory usage after sufficient uptime (Vista sidebar gadgets are first to spring to mind, a lot of them just keep using more RAM over time, quite a lot of games have done in the past also)
I liken this to stability testing of overclocked cpu's, serious overclockers don't just run their average applications to test stability, they hammer it 100% with prime95 overnight or sometimes for days to test stability under the worst conditions possible, if it fails is disregarded as a stable overclock. What i'd consider "good" testing of the stability and problems with only using RAM would be to make sure it's filled and lots of data swapping is done while it's filled. The test ran which filled the RAM caused errors and shouldn't be ignored.
The other thing you have to consider is both badly programmed applications/games and ones programmed to specifically detect the pagefile. I happen to know a game called S.T.A.L.K.E.R won't run if you dont have a pagefile enabled. Does it really need one? I don't know, I couldn't possibly say, I didn't write windows or S.T.A.L.K.E.R.
But that doesn't change the fact that you cannot play the game without a pagefile...
Now what was said before by several people is that these issues may be rare but when they crop up they increase the fault finding time significantly, which I'd have to agree with. I never touch my pagefile because I've never really seen any well documented evidence that doing so help in any noticeable way, however I have a friend who always turns his off.
Now...he grabbed a cracked copy of the gamer S.T.A.L.K.E.R and found that it wouldn't run and gave him no error in the process, he spent hours trying to get it to work before failing and giving up.
I just happened out of pure interest to read the release notes (in the nfo) on a cracked copy of the games executable from another cracking group who had "propered" the release (a "proper" is generaly where a group messes up a release and another group releases a correct or "proper" copy)
The listed reason for nuking the release was because the byte hacking that had been done in the exe to circumvent the copy protection stopped the pagefile error message from occuring if you didn't have a pagefile. so naturally I passed on the info and he replaced the exe and low and behold the error message occured this time as intended, he then had to re-enable the pagefile to play.
I'd just like to strongly emphisise a few points here.
1) The games behaviour is to not run when the pagefile is not present (irrelevant of piracy)
2) No my friend is not me, in some awkward attempt to make myself look innocent, I pirated the game too I just don't disable my pagefile so I never had the error.
3) Please realise that it wasn't piracy that caused this problem, it simply made the fix more obscure.
It's a pretty obscure turn of events, it's highly unlikely he would have found the cause of the game not working had I not happened to take a look at the "proper" notes for the new crack. I'm not describing this situation because it's a particuarly good example of the pagefile casuing errors, I'm using it to show the problems behind troubleshooting errors when the pagefile is at fault, it can be a real pain in the arse.
Anyhow, there's a good example of software not running with no pagefile, you can aquire the game and try for yourself if you wish, as far as I'm aware theres no officialy listed requirement on the game or any of its media that says it needs a pagefile so it's a bit of a minefield when getting it, you have no heads up that its not going to work.
While there remains no real benefit to disabling the pagefile, and there remains risks and problems for doing so, I will keep mine off.
I'd say if you're going to state anything as fact like dummy pagefiles are being created in certain circumstance, to list those circumstances and all the requirements we need to test and confirm you're right.
I'd also say that testing something for only a week with only average use is unlikely to show any problems, it's always going to be an extreme circumstance that knackers you if such a thing occurs, and your testing simply isn't covering that, maybe applications have memory leaks and would reach high memory usage after sufficient uptime (Vista sidebar gadgets are first to spring to mind, a lot of them just keep using more RAM over time, quite a lot of games have done in the past also)
I liken this to stability testing of overclocked cpu's, serious overclockers don't just run their average applications to test stability, they hammer it 100% with prime95 overnight or sometimes for days to test stability under the worst conditions possible, if it fails is disregarded as a stable overclock. What i'd consider "good" testing of the stability and problems with only using RAM would be to make sure it's filled and lots of data swapping is done while it's filled. The test ran which filled the RAM caused errors and shouldn't be ignored.
The other thing you have to consider is both badly programmed applications/games and ones programmed to specifically detect the pagefile. I happen to know a game called S.T.A.L.K.E.R won't run if you dont have a pagefile enabled. Does it really need one? I don't know, I couldn't possibly say, I didn't write windows or S.T.A.L.K.E.R.
But that doesn't change the fact that you cannot play the game without a pagefile...
Now what was said before by several people is that these issues may be rare but when they crop up they increase the fault finding time significantly, which I'd have to agree with. I never touch my pagefile because I've never really seen any well documented evidence that doing so help in any noticeable way, however I have a friend who always turns his off.
Now...he grabbed a cracked copy of the gamer S.T.A.L.K.E.R and found that it wouldn't run and gave him no error in the process, he spent hours trying to get it to work before failing and giving up.
I just happened out of pure interest to read the release notes (in the nfo) on a cracked copy of the games executable from another cracking group who had "propered" the release (a "proper" is generaly where a group messes up a release and another group releases a correct or "proper" copy)
The listed reason for nuking the release was because the byte hacking that had been done in the exe to circumvent the copy protection stopped the pagefile error message from occuring if you didn't have a pagefile. so naturally I passed on the info and he replaced the exe and low and behold the error message occured this time as intended, he then had to re-enable the pagefile to play.
I'd just like to strongly emphisise a few points here.
1) The games behaviour is to not run when the pagefile is not present (irrelevant of piracy)
2) No my friend is not me, in some awkward attempt to make myself look innocent, I pirated the game too I just don't disable my pagefile so I never had the error.
3) Please realise that it wasn't piracy that caused this problem, it simply made the fix more obscure.
It's a pretty obscure turn of events, it's highly unlikely he would have found the cause of the game not working had I not happened to take a look at the "proper" notes for the new crack. I'm not describing this situation because it's a particuarly good example of the pagefile casuing errors, I'm using it to show the problems behind troubleshooting errors when the pagefile is at fault, it can be a real pain in the arse.
Anyhow, there's a good example of software not running with no pagefile, you can aquire the game and try for yourself if you wish, as far as I'm aware theres no officialy listed requirement on the game or any of its media that says it needs a pagefile so it's a bit of a minefield when getting it, you have no heads up that its not going to work.
While there remains no real benefit to disabling the pagefile, and there remains risks and problems for doing so, I will keep mine off.
Just trying to prevent users from bricking a box. When I worked support had way to many hard luck stories; some of which involved, I thought my computer was too slow, so I disabled the page file... I won't tell people they can't do it, I just advise against it.
This posting is provided "AS IS" with no warranties, and confers no rights.
Archaea
[H]F Junkie
- Joined
- Oct 19, 2004
- Messages
- 11,826
so exactly how can you brick a box by disabling the page file?
You could potentially crash it. The chances of bricking the hardware are nill, and the chances of corrupting a file in the OS with NTFS on a memory crash are pretty darn minimal.
I think he meant brick the OS. I did it before. I thought I had plenty of memory, and I totally disabled the page file on a box that had 2GB of memory at that time. The OS would start up to the login screen, then when I tried to login, would state that there was no page file and then the computer would reboot. Not even a repair install would help, so I had to do a clean install.
This was about 8 years back with Windows 2k. Things may have changed, but that left a sour taste. Never again will I disable a page file.
TechieSooner
Supreme [H]ardness
- Joined
- Nov 7, 2007
- Messages
- 7,601
Exactly right.
The chances of that happening with Vista are slim though. You unplugged Windows 98 without properly shutting that thing down (or heck, even ME!!!!!!!!!!!!!!!!!)... you'd be lucky if that sucker booted again. Vista doesn't have as many of those issues.
A friend of mine knows some folks at Microsoft and said Windows Vista stores most of itself in RAM now. Whereas 98 worked off the hard drive so much (And failure to write file back correctly= crash), Windows Vista can survive it, because since it works out of RAM, it re-loads everything from disk again and you're fine.
Much like *nix does (This is why you can upgrade *nix and rarely need to reboot it). Obviously Microsoft isn't the whole way there yet.
Obviously a "friend of mine" isn't official source by a long shot, but the decreased need I see of reboots and ability to recover itself should I pull the plug, I think there is some truth to it.
Yes, that's what I meant. Sorry for the confusion.
As to vista running more in memory, I'm assuming he means SuperFetch, but that shouldn't increase stability without a pagefile.
This posting is provided "AS IS" with no warranties, and confers no rights.
Archaea
[H]F Junkie
- Joined
- Oct 19, 2004
- Messages
- 11,826
NTFS vs. FAT32
legendz411
2[H]4U
- Joined
- Jul 3, 2006
- Messages
- 3,534
Whats wrong with indexing? Does it really slow anything down after its done indexing everything?Except indexing and Windows Defender.
And whats wrong wiht windows defender? I knwo theres probably better software out there but I figured that defender did its job atleats half decently.
(Im not trying to pick a fight, I really dont know whats wrong with these services lol)
and UAC
That depends on the user. For soemone liek us, ya sure UAC is annoying, but I know othe rpeople that arnt nearly as computer literate who have been saved form installing bad stuff ont here computer because UAC makes them re-think every install they accepted lol.
So far the amount I have had to fix my sister in-laws computer becuase of dumb stuff she has done has been cut down LOTS since she switched to vista.
Except indexing and Windows Defender.
What's wrong with Windows Defender, since that is something I work on?
This posting is provided "AS IS" with no warranties, and confers no rights.
Joe Average
Ad Blocker - Banned
- Joined
- Apr 6, 2008
- Messages
- 15,459
Good lord, people are still going on about this?
Indexing slows things down unnecessarily. I'm not spending half of my time on the computer searching for something, and when I do need to search for something, I've got a pretty good idea on where it is and how to find it. Why spend so many resources (CPU time, memory, HDD thrashing) on indexing when you could spend all that on Superfetch, which in itself, will thrash your HDD memory and CPU a lot all on its own?
And about Windows Defender, I just simply choose to use Spybot, that's all. My preference, as per my last post.
and UAC
Except indexing and Windows Defender.
Well indexing, Defender, and UAC ... That is one thing, but I learned my lesson the [H]ardWay when I started messing with things like page filing ect.