If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#21
|
|||
|
|||
Page file
GT wrote
Rod Speed wrote kony wrote Alex Mizrahi wrote No, pages allocated but never used will not matter, a small amount of paging does NOT speedup anything. The system either has enough (real) memory that it doesn't have to page out (which is ALWAYS faster "IF" the use allows, if it doesn't require too much memory), it might be true if i run a single application cosuming gigabytes of RAM, but usage patterns on modern desktop computers are not like that. Wrong. Usage patterns do not change it and multiple apps merely have additive memory load which is still weighed against what the physical memory can support, or can't. If there is enough physical memory, enabling a pagefile will ALWAYS, no matter what other variable, be slower. Wrong with an OS that has enough of a clue to not use it unless its necessary because there isnt enough physical memory. True, but windows does not fall in the category you describe - it automatically loads itself into RAM, then swaps out parts to the swap file without question. It isnt that black and white. Most obviously when there is no swap file. And even with one that isnt smart enough to always do that, you havent established that what minimal use of the pagefile it does do when there is enough physical memory does slow things down any anyway, most obviously when what minimal use it does make of the page file is in the background etc. In the control panel you can set the minimum and maximum size for the swapfile. These settings do indicate to windows the smallest amount to use if the RAM is not full, it simply says the minimum size that the file will be. This avoids the swapfile becoming fragmented across the disk. Before setting the virtual memory size, I would recommend turning off virtual memory completely, rebooting, defragmenting the hard disk, then setting the minimum and maximum values to the same amount, so that the swapfile is configured in one large contiguous block on the disk and will never grow and shrink and therefore cannot become defragmented. Of course, this is my advice to someone with less RAM than the applications they use can ever need. If you have enough RAM - turn it off and things will be faster, quieter and more responsive. There's real downsides with no swap file with Win. For the (nonspecific and thus unprovable) "usage patterns on modern desktop", it could easily be that there is not enough physical memory - hence why a pagefile is enabled by default. Only the user can determine that it isn't needed. Wrong, the OS obviously can. Wrong, some OS's can, but windows doesn't. I said CAN, not DOES. i have lots of applications running (74 processes for 2 users), and i don't need all the processes all the time, certainly. i'm even running two OSes simultaneously -- Linux in vmware, but i'm working with that Linux from time to time. so, i think they'll better be swapped out. IF you don't have enough physical memory to accomdate all the rest without swapping, yes. That's not same as having enough and paging it out for no good reason. Pity about when the pagefile is just used to provide faster access to the read only files on the hard drive. Who told you that gem?!? You just restated that yourself in different words at tha top. Concentrate on the use of the word WHEN. as i've said, if some active process will need more RAM, or if some file operations will need be cached, Windows will swap out that processes anyway -- but it will swap out it's DLL and EXE pages if it cannot swap allocated memory to pagefile. RAM is just a cache for data -- some data is backed by files (executable or filemappings), some is backed by pagefiles, and some will be not backed by anything. OS might optimize better when it has flexibility to swap out some allocated memory that is not used to pagefile. certainly, OS might be wrong in it's optimizations, so it's questionable.. if you disable pagefile, you give priority to data that is explicitly allocated by applications that is not backed by anything -- so it's not swapped even if it's not used because there's no place for it. at same time some data that is more-or-less actively used -- for example, file cache that caches filesystem structure MFT -- can be swapped out. It is true that some data is needed later, or more frequently, and some isn't. That does not change the fact that swapping out data is going to be slower if there was no other use for remaining physical memory. Not if that is done in the background and its never used from the file. But loading from virtual memory IS using something from a file - its just not in the original place on the hard disk! The word IF is there for a reason. You are arbitrarily presuming there would be, and indeed sometimes there is, but that is not the same as a random idea about it being faster to page out without the specific situation of having insufficient physical memory to hold it all. You are arbitrarily presuming that an OS which does minimally use a page file when there is enough physical ram, that that minimal use of the page file has any effect on the speed of ops. No, we are stating a fact - swapping files between physical RAM and the slower hard disk (virtual memory) has a performance impact on the system. It isnt a fact at all IF that only happens in the background with the performance impact. Also, like I said above - the minimum setting for the swapfile is not an indication of how much windows will use as a minimum, just the smallest size the swapfile is allowed to be. Duh. You dont know that. Yes we do. No you dont. i have 2 GB of RAM and have some small page files. i believe it's more optimal, but i can't be sure.. If the total amount of allocated memory is beyond 2GB, yes it is more optimal. If the total amount is below 2GB, it may depend on how much of a benefit you would see from having a larger filecache (IF you adjust Windows memory management to have one, this is not a default installation condition), it is quite possible the larger filecache reduces rereads from HDD, more than the I/O to HDD from slight pagefile use. In the end, the goal is still the same- based on the specific uses of the system, to minimize access to the HDD. it's very unlinkely for windows to crash. memory allocation just fails, and actually application can handle this gracefully. btw there's one more reason to keep page file size at minimum -- some applications erroneosly allocate tons of RAM, and with large pagefile swapping make system non-responsible. with less or no pagefile, those application will simply honestly report failure.. If you're going to have a pagefile active, it should be large enough to handle the entire memory allocation from applications. Wrong. Perhaps you could stretch to more than 1 word in your reasoning here I did. - the standard is for virtual memory to be 100%-150% the size of the RAM That is just plain silly, most obviously when you go from say 1G to 2G of physical ram, with the same machine use, why would you need to double the swap file size in that situation ? The use of the swap file will DROP. why? Because that's the whole point of a pagefile, Nope Yup Nope. to virtually provide memory that's not there, It isnt the ENTIRE MEMORY ALLOCATION FROM THE APPS that matters, its the excess over the physical memory that matters. not to cause the app to not have enough even WITH it. Allocation that isn't used is not a bad thing, It can be speed wise if that page file space has to be allocated and is never used. In spades when you have chosen to run without a page file. it was that it used the HDD at all that causes the significant performance penalty, ONLY if that allocated memory is ever actually used. You proove our point Nope. - it is used all the time by Windows, whether it needs it or not! Pity that doesnt necessarily produce a performance impact when you have large amounts of physical ram, enough for whatever is run. |
#22
|
|||
|
|||
Page file
kony wrote
Rod Speed wrote Wrong with an OS that has enough of a clue to not use it unless its necessary because there isnt enough physical memory. I take it you either never used WINDOWS or never bothered to monitor pagefile activity. More fool you, as always. I used the words AN OS there for a reason. Pity about when the pagefile is just used to provide faster access to the read only files on the hard drive. That's wrong enough I don't know where to start. We'll see. Pagefile isn't necessarily any faster than rest of HDD, Wrong. Random access to the indexed page file is a lot quicker than going thru the file system with stuff that is too big for the file system cache. the only gain there is if paging unneeded things left more memory available for a (larger) filecache in physical memory. Wrong again. If you're going to have a pagefile active, it should be large enough to handle the entire memory allocation from applications. Wrong. why? Because that's the whole point of a pagefile, Nope. to virtually provide memory that's not there, It isnt the ENTIRE MEMORY ALLOCATION FROM THE APPS that matters, its the excess over the physical memory that matters. Never suggested it wasn't. He clearly did. The fact remains that it is possible a use(s) won't have any excess over physical memory but windows will still have pagefile activity. But that doesnt necessarily impact performance. |
#23
|
|||
|
|||
Page file
GT wrote
Rod Speed wrote Jaap Telschouw wrote I have Windows XP Pro and two gigs of RAM. Do I still need a page file? Yes, its safer to have one which doesnt get used much at all. Agreed - it is safer, but unfortunately windows will use it if it is there, whether it needs it or not, Yep. so your system will see a performance reduction,. Not necessarily. I suspect that any PC with 2GB of RAM is probably powerful enough to absorb the performance hit and the user won't even notice in day to day use. Or you dont understand how Win uses the page file in that situation. Its only noticable if you use some application that requires disk access and is impeeded by virtual memory activity. Wrong again, it wont be doing that enough to matter with 2G of physical ram with most normal machine uses. XP can be a little stupid with no page file at all. Not in my experience over the last 2 years Doing a tiny subset of machine use. - XP works much faster and more efficiently with it turned off. Bull**** on that much claim. |
#24
|
|||
|
|||
Page file
"Rod Speed" wrote in message
... GT wrote Rod Speed wrote kony wrote Alex Mizrahi wrote No, pages allocated but never used will not matter, a small amount of paging does NOT speedup anything. The system either has enough (real) memory that it doesn't have to page out (which is ALWAYS faster "IF" the use allows, if it doesn't require too much memory), it might be true if i run a single application cosuming gigabytes of RAM, but usage patterns on modern desktop computers are not like that. Wrong. Usage patterns do not change it and multiple apps merely have additive memory load which is still weighed against what the physical memory can support, or can't. If there is enough physical memory, enabling a pagefile will ALWAYS, no matter what other variable, be slower. Wrong with an OS that has enough of a clue to not use it unless its necessary because there isnt enough physical memory. True, but windows does not fall in the category you describe - it automatically loads itself into RAM, then swaps out parts to the swap file without question. It isnt that black and white. Most obviously when there is no swap file. And even with one that isnt smart enough to always do that, you havent established that what minimal use of the pagefile it does do when there is enough physical memory does slow things down any anyway, most obviously when what minimal use it does make of the page file is in the background etc. In the control panel you can set the minimum and maximum size for the swapfile. These settings do indicate to windows the smallest amount to use if the RAM is not full, it simply says the minimum size that the file will be. This avoids the swapfile becoming fragmented across the disk. Before setting the virtual memory size, I would recommend turning off virtual memory completely, rebooting, defragmenting the hard disk, then setting the minimum and maximum values to the same amount, so that the swapfile is configured in one large contiguous block on the disk and will never grow and shrink and therefore cannot become defragmented. Of course, this is my advice to someone with less RAM than the applications they use can ever need. If you have enough RAM - turn it off and things will be faster, quieter and more responsive. There's real downsides with no swap file with Win. Elaborate please - We have already covered the point of running out of RAM with swapfile turned off, so what downsides are you aware of with no swap file with Win? For the (nonspecific and thus unprovable) "usage patterns on modern desktop", it could easily be that there is not enough physical memory - hence why a pagefile is enabled by default. Only the user can determine that it isn't needed. Wrong, the OS obviously can. Wrong, some OS's can, but windows doesn't. I said CAN, not DOES. I'm sure some OSs can, but MS Windows CAN'T and its the Windows swapfile we are talking about so why did you bother to comment on other OSs? i have lots of applications running (74 processes for 2 users), and i don't need all the processes all the time, certainly. i'm even running two OSes simultaneously -- Linux in vmware, but i'm working with that Linux from time to time. so, i think they'll better be swapped out. IF you don't have enough physical memory to accomdate all the rest without swapping, yes. That's not same as having enough and paging it out for no good reason. Pity about when the pagefile is just used to provide faster access to the read only files on the hard drive. Who told you that gem?!? You just restated that yourself in different words at tha top. I have never stated that the swap file is used to provide faster access to read only files - this is some tripe that only you have mentioned! Concentrate on the use of the word WHEN. If I concentrate on your word WHEN, does your argument make more sense. Hold on while I try.... Nope you're still wrong. as i've said, if some active process will need more RAM, or if some file operations will need be cached, Windows will swap out that processes anyway -- but it will swap out it's DLL and EXE pages if it cannot swap allocated memory to pagefile. RAM is just a cache for data -- some data is backed by files (executable or filemappings), some is backed by pagefiles, and some will be not backed by anything. OS might optimize better when it has flexibility to swap out some allocated memory that is not used to pagefile. certainly, OS might be wrong in it's optimizations, so it's questionable.. if you disable pagefile, you give priority to data that is explicitly allocated by applications that is not backed by anything -- so it's not swapped even if it's not used because there's no place for it. at same time some data that is more-or-less actively used -- for example, file cache that caches filesystem structure MFT -- can be swapped out. It is true that some data is needed later, or more frequently, and some isn't. That does not change the fact that swapping out data is going to be slower if there was no other use for remaining physical memory. Not if that is done in the background and its never used from the file. But loading from virtual memory IS using something from a file - its just not in the original place on the hard disk! The word IF is there for a reason. There may be the occasional chunk of data that is swapped out to virtual memory, then never loaded, but there is significantly more data that is swapped out to Virtual memory and IS loaded again. Whilst swapping out data in the background is fine and won't impact performance, the performance hit is incurred when the majority of data is swapped out while you are doing something or has to be swapped back in when you need something, or need the space in the swapfile. You are arbitrarily presuming there would be, and indeed sometimes there is, but that is not the same as a random idea about it being faster to page out without the specific situation of having insufficient physical memory to hold it all. You are arbitrarily presuming that an OS which does minimally use a page file when there is enough physical ram, that that minimal use of the page file has any effect on the speed of ops. No, we are stating a fact - swapping files between physical RAM and the slower hard disk (virtual memory) has a performance impact on the system. It isnt a fact at all IF that only happens in the background with the performance impact. And as we have already discussed, swapping back in doesn't happen in the background and swapping out doesn't always happen in the background, so like I said there is a performance impact on the system. Also, like I said above - the minimum setting for the swapfile is not an indication of how much windows will use as a minimum, just the smallest size the swapfile is allowed to be. Duh. You dont know that. Yes we do. No you dont. i have 2 GB of RAM and have some small page files. i believe it's more optimal, but i can't be sure.. If the total amount of allocated memory is beyond 2GB, yes it is more optimal. If the total amount is below 2GB, it may depend on how much of a benefit you would see from having a larger filecache (IF you adjust Windows memory management to have one, this is not a default installation condition), it is quite possible the larger filecache reduces rereads from HDD, more than the I/O to HDD from slight pagefile use. In the end, the goal is still the same- based on the specific uses of the system, to minimize access to the HDD. it's very unlinkely for windows to crash. memory allocation just fails, and actually application can handle this gracefully. btw there's one more reason to keep page file size at minimum -- some applications erroneosly allocate tons of RAM, and with large pagefile swapping make system non-responsible. with less or no pagefile, those application will simply honestly report failure.. If you're going to have a pagefile active, it should be large enough to handle the entire memory allocation from applications. Wrong. Perhaps you could stretch to more than 1 word in your reasoning here I did. Well you only posted the word 'Wrong' in that discussion - perhaps you wrote the other words of that sentence in white text? Please focus on the word HERE - the standard is for virtual memory to be 100%-150% the size of the RAM That is just plain silly, most obviously when you go from say 1G to 2G of physical ram, with the same machine use, why would you need to double the swap file size in that situation ? The use of the swap file will DROP. Almost - the use of the swapfile will remain exactly the same or even increase because Windows uses it whether it needs to or not, hence our recommendation that it is turned off on a PC with sufficient RAM. However, if you insist on having a swapfile, then I repeat, "the STANDARD is for virtual memory to be 100%-150% the size of the RAM" why? Because that's the whole point of a pagefile, Nope Yup Nope. to virtually provide memory that's not there, It isnt the ENTIRE MEMORY ALLOCATION FROM THE APPS that matters, its the excess over the physical memory that matters. not to cause the app to not have enough even WITH it. Allocation that isn't used is not a bad thing, It can be speed wise if that page file space has to be allocated and is never used. In spades when you have chosen to run without a page file. it was that it used the HDD at all that causes the significant performance penalty, ONLY if that allocated memory is ever actually used. You proove our point Nope. Yes it does - you agreed that there is only a performance hit if the virtual memory is used. We have already told you that windows uses a pagefile whether it needs to or not, therefore will require to swap pages back into RAM when required, therefore there is a performance hit. So you DID prove our point. - it is used all the time by Windows, whether it needs it or not! Pity that doesnt necessarily produce a performance impact when you have large amounts of physical ram, enough for whatever is run. Let me summarise: It is a fact that reading a page from virtual memory will be slower than reading a page from physical RAM. It is a fact that Windows will use a swapfile if one is present, regardless of the amount of physical RAM in the system. It is a fact that swapping out to virtual memory while the user is working will have an impact on performance It is a fact that swapping out to virtual memory in the background will NOT affect the performance from the user's point of view It is a fact that if any page required by the OS or an application is in virtual memory, then the performance of that particular operation will be slowed than if the page had been in physical RAM. With no swap file, everything loaded will always be in physical RAM and accessed at maximum speed. With a swap file, not everything loaded will always be in physical RAM so there will be an innevitable performance reduction, which will be measurable when pages are swapped back into physical RAM and measurable when pages are swapped into virtual memory concurrently with other user activity. So with the swapfile turned on, windows will use it and the performance of the PC will be slightly reduced. With the swapfile turned off, windows cannot use it, so this particular bottleneck in performance will be elliminated, but we introduce the possibility of windows running out of memory, which will cause a runtime error in the application in question and may take the OS down with it. |
#25
|
|||
|
|||
Page file
"Rod Speed" wrote in message
... GT wrote Rod Speed wrote Jaap Telschouw wrote I have Windows XP Pro and two gigs of RAM. Do I still need a page file? Yes, its safer to have one which doesnt get used much at all. Agreed - it is safer, but unfortunately windows will use it if it is there, whether it needs it or not, Yep. Ah, so one point is agreed - windows uses a swapfile whether it need it or not so your system will see a performance reduction,. Not necessarily. But you just agreed that windows uses the swapfile whether it needs to or not, so when a swapped out file is required, it must be paged-in from the swap file with an associated delay. I suspect that any PC with 2GB of RAM is probably powerful enough to absorb the performance hit and the user won't even notice in day to day use. Or you dont understand how Win uses the page file in that situation. Careful use of the word OR there - I like it! A fair alternative, but the original statement (the bit before the or) is correct. Its only noticable if you use some application that requires disk access and is impeeded by virtual memory activity. Wrong again, it wont be doing that enough to matter with 2G of physical ram with most normal machine uses. XP can be a little stupid with no page file at all. Not in my experience over the last 2 years Doing a tiny subset of machine use. Oh you know what I do then? Clearly a lot more than you! - XP works much faster and more efficiently with it turned off. Bull**** on that much claim. Nice argument - you win! |
#26
|
|||
|
|||
Page file
On Sat, 20 Jan 2007 04:10:05 +1100, "Rod Speed"
wrote: Pagefile isn't necessarily any faster than rest of HDD, Wrong. Random access to the indexed page file is a lot quicker than going thru the file system with stuff that is too big for the file system cache. When did you plan on having random access to these giant files? Are you ignoring that rewriting these big files to a pagefile is in itself a performance loss before even considering the performance of rereads? The fact remains that it is possible a use(s) won't have any excess over physical memory but windows will still have pagefile activity. But that doesnt necessarily impact performance. Not necessarily, but usually it does, particularly when a system has only 1 hard drive. Windows can make use of idle time but at best it can only guess what to do then, and has to correct when you're actively using the system. |
#27
|
|||
|
|||
Page file
GT wrote
Rod Speed wrote GT wrote Rod Speed wrote kony wrote Alex Mizrahi wrote No, pages allocated but never used will not matter, a small amount of paging does NOT speedup anything. The system either has enough (real) memory that it doesn't have to page out (which is ALWAYS faster "IF" the use allows, if it doesn't require too much memory), it might be true if i run a single application cosuming gigabytes of RAM, but usage patterns on modern desktop computers are not like that. Wrong. Usage patterns do not change it and multiple apps merely have additive memory load which is still weighed against what the physical memory can support, or can't. If there is enough physical memory, enabling a pagefile will ALWAYS, no matter what other variable, be slower. Wrong with an OS that has enough of a clue to not use it unless its necessary because there isnt enough physical memory. True, but windows does not fall in the category you describe - it automatically loads itself into RAM, then swaps out parts to the swap file without question. It isnt that black and white. Most obviously when there is no swap file. And even with one that isnt smart enough to always do that, you havent established that what minimal use of the pagefile it does do when there is enough physical memory does slow things down any anyway, most obviously when what minimal use it does make of the page file is in the background etc. In the control panel you can set the minimum and maximum size for the swapfile. These settings do indicate to windows the smallest amount to use if the RAM is not full, it simply says the minimum size that the file will be. This avoids the swapfile becoming fragmented across the disk. Before setting the virtual memory size, I would recommend turning off virtual memory completely, rebooting, defragmenting the hard disk, then setting the minimum and maximum values to the same amount, so that the swapfile is configured in one large contiguous block on the disk and will never grow and shrink and therefore cannot become defragmented. Of course, this is my advice to someone with less RAM than the applications they use can ever need. If you have enough RAM - turn it off and things will be faster, quieter and more responsive. There's real downsides with no swap file with Win. Elaborate please - We have already covered the point of running out of RAM with swapfile turned off, so what downsides are you aware of with no swap file with Win? Thats the real downside, Win doesnt handle that well. For the (nonspecific and thus unprovable) "usage patterns on modern desktop", it could easily be that there is not enough physical memory - hence why a pagefile is enabled by default. Only the user can determine that it isn't needed. Wrong, the OS obviously can. Wrong, some OS's can, but windows doesn't. I said CAN, not DOES. I'm sure some OSs can, but MS Windows CAN'T I didnt say it does. and its the Windows swapfile we are talking about Nope, its swap files in general that are being discussed. so why did you bother to comment on other OSs? Its swap files in general that are being discussed. i have lots of applications running (74 processes for 2 users), and i don't need all the processes all the time, certainly. i'm even running two OSes simultaneously -- Linux in vmware, but i'm working with that Linux from time to time. so, i think they'll better be swapped out. IF you don't have enough physical memory to accomdate all the rest without swapping, yes. That's not same as having enough and paging it out for no good reason. Pity about when the pagefile is just used to provide faster access to the read only files on the hard drive. Who told you that gem?!? You just restated that yourself in different words at tha top. I have never stated that the swap file is used to provide faster access to read only files You did however say that Win does load some read only files into the swap file. - this is some tripe that only you have mentioned! It isnt tripe, its fact. The access to that stuff is faster from the swap file instead of to the hard drive when it isnt in the file cache. Concentrate on the use of the word WHEN. If I concentrate on your word WHEN, does your argument make more sense. Hold on while I try.... Nope you're still wrong. Nope. as i've said, if some active process will need more RAM, or if some file operations will need be cached, Windows will swap out that processes anyway -- but it will swap out it's DLL and EXE pages if it cannot swap allocated memory to pagefile. RAM is just a cache for data -- some data is backed by files (executable or filemappings), some is backed by pagefiles, and some will be not backed by anything. OS might optimize better when it has flexibility to swap out some allocated memory that is not used to pagefile. certainly, OS might be wrong in it's optimizations, so it's questionable.. if you disable pagefile, you give priority to data that is explicitly allocated by applications that is not backed by anything -- so it's not swapped even if it's not used because there's no place for it. at same time some data that is more-or-less actively used -- for example, file cache that caches filesystem structure MFT -- can be swapped out. It is true that some data is needed later, or more frequently, and some isn't. That does not change the fact that swapping out data is going to be slower if there was no other use for remaining physical memory. Not if that is done in the background and its never used from the file. But loading from virtual memory IS using something from a file - its just not in the original place on the hard disk! The word IF is there for a reason. There may be the occasional chunk of data that is swapped out to virtual memory, then never loaded, Its a hell of a lot more than an occasional chunk. but there is significantly more data that is swapped out to Virtual memory and IS loaded again. Depends entirely on how the system is used and how much physical ram there is, even with Win. Whilst swapping out data in the background is fine and won't impact performance, the performance hit is incurred when the majority of data is swapped out while you are doing something That doesnt happen when you have lots of physical ram. or has to be swapped back in when you need something, That doesnt happen when you have lots of physical ram either. or need the space in the swapfile. Doesnt happen with lots of physical ram. You are arbitrarily presuming there would be, and indeed sometimes there is, but that is not the same as a random idea about it being faster to page out without the specific situation of having insufficient physical memory to hold it all. You are arbitrarily presuming that an OS which does minimally use a page file when there is enough physical ram, that that minimal use of the page file has any effect on the speed of ops. No, we are stating a fact - swapping files between physical RAM and the slower hard disk (virtual memory) has a performance impact on the system. It isnt a fact at all IF that only happens in the background with the performance impact. And as we have already discussed, swapping back in doesn't happen in the background Swapping back doesnt happen if you can run without a swap file and have one anyway. and swapping out doesn't always happen in the background, It does in the situation where you have enough physical ram to be able to do without a swap file, and have one anyway. so like I said there is a performance impact on the system. You are just plain wrong when you have enough physical ram to be able to do without a swap file, and have one anyway. Also, like I said above - the minimum setting for the swapfile is not an indication of how much windows will use as a minimum, just the smallest size the swapfile is allowed to be. Duh. You dont know that. Yes we do. No you dont. i have 2 GB of RAM and have some small page files. i believe it's more optimal, but i can't be sure.. If the total amount of allocated memory is beyond 2GB, yes it is more optimal. If the total amount is below 2GB, it may depend on how much of a benefit you would see from having a larger filecache (IF you adjust Windows memory management to have one, this is not a default installation condition), it is quite possible the larger filecache reduces rereads from HDD, more than the I/O to HDD from slight pagefile use. In the end, the goal is still the same- based on the specific uses of the system, to minimize access to the HDD. it's very unlinkely for windows to crash. memory allocation just fails, and actually application can handle this gracefully. btw there's one more reason to keep page file size at minimum -- some applications erroneosly allocate tons of RAM, and with large pagefile swapping make system non-responsible. with less or no pagefile, those application will simply honestly report failure.. If you're going to have a pagefile active, it should be large enough to handle the entire memory allocation from applications. Wrong. Perhaps you could stretch to more than 1 word in your reasoning here I did. Well you only posted the word 'Wrong' in that discussion Wrong. The rest is still in the quoting below. - perhaps you wrote the other words of that sentence in white text? You'll end up blind if you dont watch out. Please focus on the word HERE Focus on the rest of the quoted text below. - the standard is for virtual memory to be 100%-150% the size of the RAM That is just plain silly, most obviously when you go from say 1G to 2G of physical ram, with the same machine use, why would you need to double the swap file size in that situation ? The use of the swap file will DROP. Almost No almost about it, your original just plain silly. - the use of the swapfile will remain exactly the same or even increase because Windows uses it whether it needs to or not, Wrong when the 1G of physical ram isnt enough for what is being done. hence our recommendation that it is turned off on a PC with sufficient RAM. Separate matter entirely to your stupid 100-150% claim with that much ram. However, if you insist on having a swapfile, then I repeat, "the STANDARD is for virtual memory to be 100%-150% the size of the RAM" Wrong with that much physical ram. why? Because that's the whole point of a pagefile, Nope Yup Nope. to virtually provide memory that's not there, It isnt the ENTIRE MEMORY ALLOCATION FROM THE APPS that matters, its the excess over the physical memory that matters. not to cause the app to not have enough even WITH it. Allocation that isn't used is not a bad thing, It can be speed wise if that page file space has to be allocated and is never used. In spades when you have chosen to run without a page file. it was that it used the HDD at all that causes the significant performance penalty, ONLY if that allocated memory is ever actually used. You proove our point Nope. Yes it does - Nope. you agreed that there is only a performance hit if the virtual memory is used. But I dont agree that there is always performance hit when you have enough physical ram to not need the swap file, when Win uses the swap file anyway IN THAT SITUATION. We have already told you that windows uses a pagefile whether it needs to or not, No news to me, I said that myself. therefore will require to swap pages back into RAM when required, Wrong when there is enough physical ram to not need to do that. therefore there is a performance hit. Nope, because that doesnt happen when you have enough physical ram so that you can go without a swap file and have one anyway. So you DID prove our point. Nope. - it is used all the time by Windows, whether it needs it or not! Pity that doesnt necessarily produce a performance impact when you have large amounts of physical ram, enough for whatever is run. Let me summarise: It is a fact that reading a page from virtual memory will be slower than reading a page from physical RAM. Not when you have enough physical ram to be able to do without a swap file and have one anyway. It is a fact that Windows will use a swapfile if one is present, regardless of the amount of physical RAM in the system. That's the only bit you did manage to get right and is no news. It is a fact that swapping out to virtual memory while the user is working will have an impact on performance Win doesnt do that when you have enough physical ram to be able to do without a swap file and have one anyway. It is a fact that swapping out to virtual memory in the background will NOT affect the performance from the user's point of view What I said. It is a fact that if any page required by the OS or an application is in virtual memory, then the performance of that particular operation will be slowed than if the page had been in physical RAM. Not when you have enough physical ram to be able to do without a swap file and have one anyway. With no swap file, everything loaded will always be in physical RAM and accessed at maximum speed. Just as true when you have enough physical ram to be able to do without a swap file and have one anyway. With a swap file, not everything loaded will always be in physical RAM Wrong. so there will be an innevitable performance reduction, Wrong. which will be measurable when pages are swapped back into physical RAM Win doesnt do that when you have enough physical ram to be able to do without a swap file and have one anyway. and measurable when pages are swapped into virtual memory concurrently with other user activity. Win doesnt do that when you have enough physical ram to be able to do without a swap file and have one anyway. So with the swapfile turned on, windows will use it Yes. and the performance of the PC will be slightly reduced. Nope. With the swapfile turned off, windows cannot use it, You quite sure you arent one of those rocket scientist pig ignorant children ? so this particular bottleneck in performance No such animal when you have enough physical ram to be able to do without a swap file and have one anyway. will be elliminated, but we introduce the possibility of windows running out of memory, which will cause a runtime error in the application in question and may take the OS down with it. So, when you have enough physical ram to be able to do without a swap file and have one anyway, that cannot happen. You can run up the white flag now. |
#28
|
|||
|
|||
Page file
kony wrote
Rod Speed wrote Pagefile isn't necessarily any faster than rest of HDD, Wrong. Random access to the indexed page file is a lot quicker than going thru the file system with stuff that is too big for the file system cache. When did you plan on having random access to these giant files? Nothing to do with random access to giant files, everything to do with what files end up not in the file system cache. Are you ignoring that rewriting these big files to a pagefile I never ever said that they ever get into the pagefile. is in itself a performance loss before even considering the performance of rereads? See above. The fact remains that it is possible a use(s) won't have any excess over physical memory but windows will still have pagefile activity. But that doesnt necessarily impact performance. Not necessarily, but usually it does, particularly when a system has only 1 hard drive. Nope, not when you have enough physical ram to be able to do without a swap file and have one anyway. Windows can make use of idle time but at best it can only guess what to do then, and has to correct when you're actively using the system. Nope, not when you have enough physical ram to be able to do without a swap file and have one anyway. |
#29
|
|||
|
|||
Page file
GT wrote
Rod Speed wrote GT wrote Rod Speed wrote Jaap Telschouw wrote I have Windows XP Pro and two gigs of RAM. Do I still need a page file? Yes, its safer to have one which doesnt get used much at all. Agreed - it is safer, but unfortunately windows will use it if it is there, whether it needs it or not, Yep. Ah, so one point is agreed - windows uses a swapfile whether it need it or not Yep, but your performance claim isnt agree when you have enough physical ram to be able to do without a swap file and have one anyway. so your system will see a performance reduction,. Not necessarily. But you just agreed that windows uses the swapfile whether it needs to or not, so when a swapped out file is required, it must be paged-in from the swap file with an associated delay. I never agreed that it was swapped out in the sense that it ever needs to be swapped in again when you have enough physical ram to be able to do without a swap file and have one anyway. In fact that doesnt happen. I suspect that any PC with 2GB of RAM is probably powerful enough to absorb the performance hit and the user won't even notice in day to day use. Or you dont understand how Win uses the page file in that situation. Careful use of the word OR there - I like it! A fair alternative, but the original statement (the bit before the or) is correct. Nope. Its only noticable if you use some application that requires disk access and is impeeded by virtual memory activity. Wrong again, it wont be doing that enough to matter with 2G of physical ram with most normal machine uses. XP can be a little stupid with no page file at all. Not in my experience over the last 2 years Doing a tiny subset of machine use. Oh you know what I do then? No one ever does more than a tiny subset of machine use. Clearly a lot more than you! Clearly not. You clearly dont actually understand how Win uses the swap file when you have enough physical ram to be able to do without a swap file and have one anyway. - XP works much faster and more efficiently with it turned off. Bull**** on that much claim. Nice argument - you win! Your pathetic excuse for an 'argument' in spades. Yours was just an assertion without a shred of evidence to substantiate it. In spades with the much claim. |
#30
|
|||
|
|||
Page file
On Sat, 20 Jan 2007 08:45:16 +1100, "Rod Speed"
wrote: kony wrote Rod Speed wrote Pagefile isn't necessarily any faster than rest of HDD, Wrong. Random access to the indexed page file is a lot quicker than going thru the file system with stuff that is too big for the file system cache. When did you plan on having random access to these giant files? Nothing to do with random access to giant files, everything to do with what files end up not in the file system cache. Still doesn't remove the performance penalty of reading it in, paging it out, and still reading it in again. If the large files are really such a problem, get more physical memory. It's not as though a pagefile is a magic fix for this either, the pagefile does not just receive any files too big for the filecache, it is not large enough to do that even if it did. Not necessarily, but usually it does, particularly when a system has only 1 hard drive. Nope, not when you have enough physical ram to be able to do without a swap file and have one anyway. Wrong, _IF_ you really do have enough physical ram to be able to do without, enabling the paging feature will only cause more work for windows and accessing a slow(er) medium, versus not doing it at all. Windows was wrote dumb to assume you don't have enough physical memory and start paging out beforehand, even if you don't ever have the need in the specific use of the system. I will write again that many people don't actually have enough physical ram, and cannot then disable the pagefile. The issue of when it has a performance gain (disabled) depends entirely on having ample (some would even consider it excess because they think in terms of reliance of a pagefile) physical ram. Windows can make use of idle time but at best it can only guess what to do then, and has to correct when you're actively using the system. Nope, not when you have enough physical ram to be able to do without a swap file and have one anyway. Doesn't matter if you have enough physical ram, if the pagefile is enabled it will page and it will be a performance loss if you didn't need the addt'l memory it freed up. Some do, some don't. |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Seagate Barracuda 160 GB IDE becomes corrupted. RMA? | Dan_Musicant | Storage (alternative) | 79 | February 28th 06 08:23 AM |
Nothing But Coasters | dadiOH | Cdr | 11 | February 28th 06 12:48 AM |
errors using retrospect express s-ware --- too many?? | MB_ | Dell Computers | 5 | October 14th 05 08:50 PM |
my dvd burner keeps having problems | nullboy | Cdr | 3 | September 9th 05 01:46 AM |
P4t533 Max Ram | John Smith | Asus Motherboards | 26 | September 7th 04 04:09 AM |