Openldap on a linux cluster

[parthasarathi s a <parthasarathi@odysseytec.com>]

Hi all,
I'm using the Openldap server to house a huge database of around 10 million entries.I've installed OpenLdap on a p3 800mhz with 20gig hard disk running Redhat Linux 7.2. It is very obvious that more the number of entries the slower the transaction. I want to know is there ary alternative. I have thought of clustering. I would like to know if we can have OpenLDAP installed on a Linux cluster. The main idea behind it would be to increase the transaction speed. I would like to do a query on a openldap server containing around 10 million entries. 
I have several questions regarding this:
	
	- Is there any way of increasing the transaction speed.
	- would clustering help????
	- would recompiling the code help? If so what are the compiler options/flags we got to use.
	- Is there *any* way by which i can get a speedy response from a ldap server housing around 10 million 		entries(upper limit).
The 10 million entries would be something like a telephone directory of a large geographical area..

_____________
parthasarathi s a
e-mail: parthasarathi at odysseytec dot com
___________________ 
`When you say " I wrote a program that crashed Windows", people just stare at you blankly and say "Hey, I got those with the system, *for free*".'
(By Linus Torvalds)
___________________