You are not logged in.
Running 2.6.4, up to date with all packages, runs 24/7.
756K memory, Celeron 766
Using 12 windows:
1 - superuser file manager
2 - superuser console
3 - kde system guard, kdiskfree
12 - opera browser, 7 tabs open
others normally used for openoffice, gimp, kmail, some others, but currently idle to examine this memory problem.
Last night memory allocations were
500,000 free and 0 swap.
The only activity overnight may have been a news site that reloads every ten minutes.
Here are the memory stats this morning:
[root@bill billb]# free
total used free shared buffers cached
Mem: 772776 762752 10024 0 294224 188488
-/+ buffers/cache: 280040 492736
Swap: 1172704 46408 1126296
What can possibly be going on here?
I have observed that given these stats a cd ripper is extremely slow and killing it leaves memory cluttered. Even shutting everything down does not reclaim free memory and dump the swap allocation.
The only way I know to restore memory is to reboot.
I found a website thru google that explores the kernel's memory management but, surely ordinary desktop users don't have to become memory gurus. I hope that someone can give me a solution for quicker use.
Even as I write this memory is being allocated.
[root@bill billb]# free
total used free shared buffers cached
Mem: 772776 764920 7856 0 294248 189256
-/+ buffers/cache: 281416 491360
Swap: 1172704 46404 1126300
Could system and kde diskfree be responsible?
Thanks for your help.
Bill Barnes
Offline
See the part that says cache? That's memory being used to cache hard drive information. It's not strictly being used and can fairly eaily be freed, but it contributes to the used memory section. To see the amount of free memory that's not disk cache look at the free column of the cache row.
I have discovered that all of mans unhappiness derives from only one source, not being able to sit quietly in a room
- Blaise Pascal
Offline
The only activity overnight may have been a news site that reloads every ten minutes.
No. You are forgetting something : the cron jobs. Every night, (almost) every Unix system runs system maintennace jobs (see crontab -l as root to find yours). One of the tasks is updating the locate database, i.e. looking through the whole hard drive.
That's what Xentac is pointing to : the disk cache. It got filled ....
Offline