Horst von Brand <vonbrand@pincoya.inf.utfsm.cl>:
> orc@pell.portland.or.us (david parsons) said:
> > In article <linux.kernel.45hadsgku4f59qae3ouohgbk7k4p6lc5os@4ax.com>,
> > James Sutherland <jas88@cam.ac.uk> wrote:
>
> > >Now we'll take a WWW server, with 100 processes forked, all sharing
> > >most of the image. You just blew 2Gb or so of my swap space, to
> > >achieve - nothing.
>
> > Okay, I'm getting really curious here: what application do you
> > have that requires that you run 100 copies of a web server each
> > with 20mb of unique writable data?
>
> 100 concurrent accesses to your site, no memory overcommit.
It doesn't take that much - of that 20MB:
the parent process occupies 2-3MB - ~2MB shared text + at most 1MB data.
Each child will vary from 3MB (unused server) to maybe 10MB - 20 if
the perl module is included. If CGI is used, then each child
will have 2MB shared text + 2-3MB data + CGI.
The CGI memory is reclaimed as they exit.
On the other hand, lets assume the amount IS accurate:
Without overcommit:
You can support 100 simultaneous connections, with full saturation of
each server, with no failures.
Result: happy customers, happy management, maybe even a raise.
If you overcommit:
You might support 100 simultaneous connections, but not full saturation
of each server, without aborting some connections or crashing the
server/system.
Result: some unhappy customers, not so happy management, difficulty
in identifying problems, and definitely no raise.
So you used 2G byte of swap - I know where you can by 18GB of swap for about
$300 US...
-------------------------------------------------------------------------
Jesse I Pollard, II
Email: pollard@navo.hpc.mil
Any opinions expressed are solely my own.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Thu Mar 23 2000 - 21:00:30 EST