have a look http://apache.ivn.cl/ may be interesting
George Barnett schrieb:
Hmm,
That makes sense to me, although it sounds like a great way to kill a
backend. Given the thread limit per processes is 64, you'd end up
having almost no way of controlling connections to the backend.
George
Christian Gottschalch wrote:
Hi,
i think it ist per process, so if you have 100 processes with max
value 200 your Hard Maximum will be 20000 connections, but im not shure.
Documentation says
(http://httpd.apache.org/docs/2.2/mod/mod_proxy.html#proxypass):
The default for a Hard Maximum for the number of connections is the
number of threads per process in the active MPM.
which makes sense, because every thread can handle one request, to
handle the request a minimum of one connection to the proxyed server
is needed, therefore i think it must be a per process setting. In
prefork mode it is set to 1, because if you use prefork only one
request can be handled per process.
regards.
George Barnett schrieb:
Hi All,
I've asked this on irc and it was suggested that I try asking the dev
because the docs aren't clear, but I'll ask here aswell.
I'm using apache httpd 2.2.3 with the event MPM to do some traffic
throttling to a backend. What I have is this:
First, traffic hit a default vhost with lots of rewrite rules such
as this:
RewriteCond %{REQUEST_URI} ^/page.*$ [NC]
RewriteCond %{QUERY_STRING} ^.*pattern.*$ [NC]
RewriteRule ^(.*)$ http://trafficclass.localhost/$1 [P,L]
In the hosts file, trafficclass.localhost is 127.0.0.1, so from
there, the
traffic heads back into a vhost.
Then I have a vhost with the following in it:
<VirtualHost 127.0.0.1:80>
DocumentRoot /opt/apache/htdocs
ServerName trafficclass.localhost
ProxyRequests off
ProxyBadHeader Ignore
CustomLog /var/log/apache2/gold-access.log combined
Errorlog /var/log/apache2/gold-error.log
<Location />
ProxyPass http://1.2.3.4/ max=200 acquire=500
# IIS Sucks
SetEnv force-proxy-request-1.0 1
SetEnv proxy-nokeepalive 1
</Location>
</Virtualhost>
In my setup I have 100 processes with 64 threads per process.
Last night during exceptional load, I saw over 3200 connections to the
backend. My question - is it possible that max=200 is a per process
and
not server wide limit?
If so, there's a way to set a server wide limit to the backend?
Cheers,
George
---------------------------------------------------------------------
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: [EMAIL PROTECTED]
" from the digest: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
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: [EMAIL PROTECTED]
" from the digest: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
