I am seeking advice on tuning MPM worker parameters for a special use.
I have 4 redundant servers in a local network , each serving about 64
clients. The servers have a custom module used to serve special data to
a particular application on the clients. The Apache server is not used
for anything else and this is a private network. Maximum usage could be
as high as 160 clients during a failure of one or more other servers.
Each client will have only one connection alive at a time with Apache
and all requests are sequential, there is no pipelining. At times all
clients will be started at once and flood the server with requests to
get the data they need for start-up. The application will continue to
request data over time at varying rates. There will also be times when
most of the applications are not running, they close their connection to
the server. I am also using mem-cache on the servers to minimize disk
access to the static data being served.
What settings do you suggest to provide for smooth start-up for 64
clients connecting at once and a maximum load of 160 clients? I am
currently seeing a more than the two initial client-servers starting up
when all 64 clients connect at once. Here are my current settings:
KeepAlive On
MaxKeepAliveRequests 0
KeepAliveTimeout 60
<IfModule mpm_worker_module>
StartServers 2
MaxClients 192
MinSpareThreads 25
MaxSpareThreads 75
ThreadsPerChild 64
MaxRequestsPerChild 0
</IfModule>
Thank you,
Chuck
---------------------------------------------------------------------
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-unsubscr...@httpd.apache.org
" from the digest: users-digest-unsubscr...@httpd.apache.org
For additional commands, e-mail: users-h...@httpd.apache.org
