You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@httpd.apache.org by Thorsten von Eicken <tv...@voneicken.com> on 2002/01/13 02:54:32 UTC
keep-alive & shedding load
We're running into resource problems with apache 1.3. We are currently
serving between 100 and 150 requests/sec. The cpu usage is below 25% (quad
sun solaris box) but we have some 600-odd apache processes running. Our
current config allows up to 1000 processes, but we have "only" 1GB of
memory and at some point soon it will start swapping. Looking at the
server-status, it seems that the majority of the processes are actually
just in the keep-alive state, which currently has a timeout of 15 seconds.
Clearly we can reduce the timeout, but is seems it would be great if apache
simply started recycling processes in keep-alive state if it runs out of
processes. Does it do that? Will 2.0 do that? I haven't had the time to dig
into the code to see whether there is any such load shedding algorithm, so
if someone knows the answer, that would be great.
Thanks,
Thorsten
-------------------------------------------------------------------------
Thorsten von Eicken
expertcity.com Phone: (805) 690-6445
5385 Hollister Ave #111 Fax: (805) 964-6103
Santa Barbara, CA 93111 Cell: (805) 403-0982
---------------------------------------------------------------------
The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:http://httpd.apache.org/userslist.html> for more info.
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org
RE: keep-alive & shedding load
Posted by Joshua Slive <jo...@slive.ca>.
> From: Thorsten von Eicken [mailto:tve@voneicken.com]
> We're running into resource problems with apache 1.3. We are currently
> serving between 100 and 150 requests/sec. The cpu usage is below
> 25% (quad
> sun solaris box) but we have some 600-odd apache processes running. Our
> current config allows up to 1000 processes, but we have "only" 1GB of
> memory and at some point soon it will start swapping. Looking at the
> server-status, it seems that the majority of the processes are actually
> just in the keep-alive state, which currently has a timeout of 15
> seconds.
> Clearly we can reduce the timeout, but is seems it would be great
> if apache
> simply started recycling processes in keep-alive state if it runs out of
> processes. Does it do that? Will 2.0 do that? I haven't had the
> time to dig
> into the code to see whether there is any such load shedding
> algorithm, so
> if someone knows the answer, that would be great.
Currently, there is nothing that you can do other that manually change the
timeout.
I think having a more automated system for handling this would be great.
You might want to suggest that on the development list.
Joshua.
---------------------------------------------------------------------
The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:http://httpd.apache.org/userslist.html> for more info.
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org