Randy Kramer escribió:

> Oscar wrote:
> > Run an application in Windows. Fast!... but... suppose the application hangs.
> > What happens to windows?...
> > Run an application in Linux. Less fast... but... suppose the application
> > hangs.
> > What happens to Linux? The security has a price ;-)
>
> You have a point, but I've used Windows for so long that I've learned
> ways to avoid or minimize crashes.

I was thinking it before using linux :-)

> > BTW, these answers are not referred to the original subject by Fred:
> > $ free
> >           total      used      free   shared   buffers  cached
> > Mem: 223360 219292 4068  236        576      115792
> > I suppose he's talking about the "cached" memory.
> > But I don't know the answer.
> > Any idea?
>
> Yes, I suppose he is, or generally the swap (I keep top running in a
> console).  I don't know of a way to reduce the amount of cached memory
> used -- I think your system will slow down if Linux wants to swap
> something out of memory and doesn't find any swap available.
>
> regards,
> Randy Kramer

-----------------

Mmm...
I don't know if we are talking about the same.
I will explain me in a better way (and excuse my english, please)

1. We have the cached memory. When the system reads a file from the disk, the data
will remain in a RAM memory area called "cache". When the system must read again
the same file, this file is gived from the RAM cache, not from disk. Then, the
file is readed more fast (since RAM is faster than hard disk). *cached memory
resides in RAM*

2. On the other hand, we have the swap memory. When the system must use more
physical memory (RAM), it will look for it in RAM, but if not more RAM is
available, the system will swap some memory pages (not in use) from physical
memory to the swap partition/file. When these memory pages are required again,
they are read from swap and located in the physical memory. *Swap memory resides
in HD*

If too memory is used for cached files, then there are less memory available to
applications, and the system must "swap" memory pages to the swap file/partition.
This operation is time-consuming.
In Windows, you can edit c:\windows\system.ini and look for the section called
[vcache]
You can put here two parameters:
MinFileCache=
MaxFileCache=
It will control the minimum and maximum size of RAM that windows uses for cache
(not for swap). It can revert in a better performance.
Since kernel 2.4.x, linux uses more RAM space for the cache, and in computers with
little memory it can suppose less performance because the system must swap memory
pages to the disk more frequently.
The original question of this mail subject is:
Are there any equivalent in Linux to the windows parameters "MinFileCache" and
"MaxFileCache"
Salu2
óscar.


--
   . ---
   |o_o |
   |:_/ |   Usuario
  //   \ \  de Linux
 (|     | ) Registrado
/'\_   _/`\ #227443
\___)=(___/ http://counter.li.org/




Want to buy your Pack or Services from MandrakeSoft? 
Go to http://www.mandrakestore.com

Reply via email to