httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dean Gaudet <dgau...@arctic.org>
Subject Re: mmap() versus read() (fwd)
Date Fri, 06 Mar 1998 20:05:15 GMT
Not portably that I'm aware of... 

The linux problem wasn't a maxclients problem though.  It's a problem
where 80Mb + 35 Mb + lots of small Mb for other tasks > 128Mb.  The 80 is
the amount of data trying to be served, the 35 is the size of Arctic MUD
and it's always in ram and busy.  The end result is that i/o has to be
done and with mmap() linux 2.0.x is less efficient at the i/o.

Dean

On Fri, 6 Mar 1998, Brian Behlendorf wrote:

> At 01:56 AM 3/6/98 -0800, Dean Gaudet wrote:
> >Everything runs on linux 2.0.x really well until you get a working set
> >larger than RAM. 
> 
> I've been thinking - would it be possible for Apache to simply not spawn
> more children than it had room for in RAM?  It couldn't be perfect of
> course since there are usually other applications running on a server, but
> instead of having to ask the person managing the config files to guess what
> MaxClients should be, couldn't we automate that somehow?  E.g. "MaxClients
> RAMLimit"  :)
> 
> 	Brian
> 
> 
> --=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=--
> specialization is for insects				  brian@organic.com
> 


Mime
View raw message