Page file vs RAM nuisance

Message boards : Number crunching : Page file vs RAM nuisance

To post messages, you must log in.

AuthorMessage
FoldingSolutions
Avatar

Send message
Joined: 2 Apr 06
Posts: 129
Credit: 3,506,690
RAC: 0
Message 41958 - Posted: 8 Jun 2007, 13:40:44 UTC

When running Rosetta, a large chunk of virtual memory is consumed when there is clearly still plenty of memory to be had. As the power consumption of a distributed computing project is taken into account by many users, as is the noise made by hard drives. Would it not make more sense to allow a work unit to fully load into memory and therefore allow the hard-drive (30W/h+) to go to sleep. As I run Rosetta from a USB memory stick there is no need for Rosetta to wake the hard drive up to checkpoint it's work either. Is this an issue that can be fixed or is it simply a problem with Windows? I have tried turning off the page file but when I re-boot Windows simply creates a new one.
ID: 41958 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1832
Credit: 119,860,059
RAC: 1,141
Message 41966 - Posted: 8 Jun 2007, 15:48:17 UTC
Last modified: 8 Jun 2007, 15:48:46 UTC

virtual memory management is entirely down to windows but nothing will be swapped out of RAM if there isn't something to swap in to replace it. if the page-file is re-creating on each reboot then something's wrong - if you turn it off then it should stay off (though control panel > system > advanced > performance > advanced > virtual memory > no paging file, and make sure you hit 'Set'!).

If you are turning the paging file off then you will run into problems if you start to run out of RAM. I've got a media centre that uses compactflash instead of a hard drive and that doesn't have a paging file, but it needs 1GB RAM or it runs out of memory. If you're running on 512MB I recommend you leave the paging file on.

also, if running from usb it's probably worth increasing the 'write to disk at most' value as they have a limited number of writes. Also, defragging it can increase the read speed quite dramatically (although you wouldn't think it would make a difference with it being solid-state!).

You can use sysinternals filemon to see what's accessing the HD and waking it up.
ID: 41966 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
FoldingSolutions
Avatar

Send message
Joined: 2 Apr 06
Posts: 129
Credit: 3,506,690
RAC: 0
Message 41967 - Posted: 8 Jun 2007, 16:03:31 UTC - in response to Message 41966.  

You can use sysinternals filemon to see what's accessing the HD and waking it up.[/quote]

That's good advice, what is sysinternals filemon and where do i find it?

ID: 41967 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dcdc

Send message
Joined: 3 Nov 05
Posts: 1832
Credit: 119,860,059
RAC: 1,141
Message 41968 - Posted: 8 Jun 2007, 16:05:46 UTC

ID: 41968 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
FoldingSolutions
Avatar

Send message
Joined: 2 Apr 06
Posts: 129
Credit: 3,506,690
RAC: 0
Message 41970 - Posted: 8 Jun 2007, 16:13:20 UTC - in response to Message 41968.  

Nice one, according to this program it seems to be my Norton Antivirus which is causing all the trouble.

ID: 41970 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Page file vs RAM nuisance



©2025 University of Washington
https://www.bakerlab.org