20 total posts
Did you use the /3GB switch?
I left out details on purpose to see if you know about it.
Thanks for Your Response
I did a search on the forum for /3GB switch, and made out that it is to remove 1GB and see if that helps. Unfortunately, my CPU is not very easily accessible. If that is what the /3GB switch is, I will try waiting for microsoft to supply a Vista software patch.
Sorry there will be no patch for this.
And it does not remove 1GB but tells the OS to allow 1GB more ram to be used by applications.
Isn't it sad they don't give out manuals or training for this anymore?
I'm pretty untalented with regard to digital. But it seems wiser to take something good and make it better instead of making something new just for sales revenue. So it is sad there is no manual.
If it is technically easy and not risky, I think I will try the change to the boot.ini file. I am looking at BCDedit procedures, now.
Found something that seems to fit the problem.
Well that's from 2007?
I'd enable the 3G first and if it's a suspect you want to discuss, reveal the antivirus and firewall you use.
Here I'm using Free AVG and Zonealarm free.
Computer Purchase Oct. 2007
Thanks; the /3GB switch is something I am thinking about.
I use Avast! antivirus and the OS firewall. I use dial-up Internet.
Doesn't task manager in Vista ...
show each processes memory usage? You might see one growing and growing.
Or it's spaced used by the OS itself, and not shown there. Or a memory leak indeed (some Firefox add-ons seem to have one) so it's just gone. You don't know until you try.
Can you relate the growing memory use to any active program? Like: it doesn't happen if I boot and just show the desktop, but only if I start using Windows Mail?
Thanks for the Thoughts, Kees
Indeed, that is exactly it. I watch the cached memory gradually increase while the free memory decreases. I really can't even say that this see-saw effect is even a real problem.
I use Internet Explorer and Windows Mail just about exclusively. When I rebooted over a week ago, the cached memory was at about 500MB while the free started at 2000MB. Before that reboot, the cached was about 2500MB while the free had dropped down to around 150MB.
Today, I woke the computer from hibernation. It had restored the free to 2000MB and reduced the cached to 500MB. Or it was something like that. I might like to try to bring it out of hibernation without using either IE or Windows Mail for a while to see what happens.
Down to a point, every time I bring it out of sleep mode, the free memory starts about 100MB lower than it was in the prior session.
The status of the system immediately BEFORE hybernation should be equal to that immediately AFTER waking up. Then hybernation shouldn't be a factor at all, and all that matters is what you do when the machine is awake. Is that the case indeed?
After hibernation: free memory restored
Last night before hibernation, I saw that the free memory was around 800MB while the cached was around 2300MB. Right after bringing it out of hibernation this morning, the free was about 2000MB while the cached was about 500MB or so.
That's unexpected (for me, anyway).
Just going into and out hybernation changing the free memory from 800 Mb to 2000 Mb. Now find out what process causes it to go from 2000 Mb to 800 Mb when in use.
By the way, why don't you turn off the machine during the night?
I don't like turning it off. It takes about five minutes to get back up again. Worse is that when it turns on, one of the motors revs up to its max, and that seems pretty detrimental.
I like standby since that seem to be the gentlest process.
My second choice is hibernation, but unfortunately, that also does the motor revving thing.
That motor rev is...
Usually the CPU fan. If you have the nicer ball bearing fans I expect these to work for about 5 years plus if you do your yearly maintenance and cleaning.
It's sad they didn't tackle "sound" in the design to not alarm owners like this.
Ahh, if it is that, it could have the intended purpose of blowing dust off the heatsink.
I expect that to happen.
But I'm an older programmer with too much time under the hood. You may want to consider that to speed up hibernation a call is made to all running programs to say "Hey, we're going into hibernation so you've been warned." Some programs are smart to eject cached objects so the time getting to hiber and back is faster.
Do you consider this a design flaw?
Could be true ...
I'd expected it to be an exact image of RAM being written to disk as it is, but some optimization certainly is possible.
Anyway, it's a 20 second solution/work-around to the whatever problem causes the increasing memory use if OP can't find it. Hibernate and immediately activate again. Most of us won't mind 20 seconds once a day.
Some programmer chatter.
"Example: Linux just saves and then restores the whole memory (including caches) when going into hybernate mode. This means up to several mins before going down (around 1min per each GB..."
So you can see there would be a big downside to saving every bit. Let's hope no one changes how it works next year.
Seems not too much me
These things have pretty many quirks about them such that I usually try to get around them as best as I can, with the least effort, though.