PURPOSE AND OPERATION:

The module performs the checks if the number of HTTP requests from a single IP 
exceeds a certain allowed maximum during a chosen period of time. If so, it
makes the HTTP server respond to the requests from such IP addresses with a
chosen HTTP response code (e.g, 402). The checks can be performed either 
globally for the whole web-server, or for certain URLs that match some pattern. 

INSTALLATION:
This module works with Apache2 verions 2.0.52 and greater and requires the 
following packages:
apr_memcache http://svn.northnitch.com/apr_memcache/trunk/
(uses the locally included code, and will not install it)

m4 'helper' modules http://svn.outoforder.cc/svn/autoconf/trunk/m4

memcached http://code.sixapart.com/svn/memcached/trunk/


To build the module, run
./configure
make
Then `su' to root and  
make install
Usually, this will do the job. See the output of ./configure --help for more
configuration options.

The sample configuration to load and set up the module is located in 
mod_mod_ip_count.conf. 

This code is licenced under ASF License, version 2.0

autoconf magic was stolen/from (and the original author didn't succeed much in 
that, so many modifications were done and there are many more to do) 
mod_transform.. a excellent module from svn.outoforder.cc 

USAGE AND CONFIGURATION DIRECTIVES
Once installed, the module needs to be configured globally (a per-virtual-host
configuration seems useless now). Here is the list of presently available
configuration directives and their usage:

MemCacheServers
A list of memcached servers the module will use. By default, a newly installed
memcached will bind to localhost:11211, and it is the default setting for the
module, so you may just skip this setting if you didn't reconfigure memcached

UriList
List of sample patterns. If the directive is set, only the requests containing
such pattern inside the URI part will be counted, and the rest will be ignored.
If the directive is omitted the module will operate on all requests. 

MemCacheMaxTime
MemCacheMaxRequests
Set the time in seconds and the number of requests per IP to serve during this
time. The requests that come after such limit is exceeded are replied with some
HTTP code instead of being processed normally. Note that a single user opening
a single webpage generates several requests (as an excelent example, see your
Apache log after directing your web browser to a simple phpinfo page). 
Therefore, you might probably have to gather some per-page request statistics
for your site to figure out the appropriate values.

MemCacheBlockTime
Optionally, set the time in seconds during which the request from IP's which
have exceeded their maximum are being blocked.

HttpResponse
Numeric HTTP return code to be sent in response to the requests after their 
number excceeds the permitted limit. If omitted, the default 402 return code
will be used

MemCacheAllow
IP-adresses and/or `env' variables with which an unlimited number of requests
is permitted

Latency
When it's impossible to tell exatly, how many requests will user generate
with a single hit on a web address, this directive may be used to count
all the requests coming during the specified period of time (in microseconds,
0,000001 second) from a single IP as a single request. The option is not
mandatory