Unfortunately even such a system [using MySQL and InnoDB tables] has limits and unfortunately replication is no good solution here to scale further as we will always need a master for writing the updated session data. By using replication we can take some load from it and we can configure a slave which can be promoted to master to keep session alive if the primary master machine fails but at some point in time we need another solution ... but, happy news, again: One doesn't have to look far as MySQL cluster will be happy to help. MySQL Cluster "is a high-availability, high-redundancy version of MySQL adapted for the distributed computing environment," as the MySQL documentation states.

He describes the setup (after pointing to this post about installing MySQL Cluster for memcache) and includes some sample code/SQL/ini settings you'll need to use to get PHP's memcached functionality to cooperate with it.

Often for my clients, I have to prepare the deployment strategy for their LAMP based web applications. Some of them are small to medium businesses and are starting up so a single server setup work out for them. But there are few large web applications too which are growing continuously in terms of users and demands scaling either horizontally or vertically.

They give a brief overview of what vertical and horizontal scaling are as well as a simple layout of a basic PHP-based cluster. They list some of the requirements for this simple cluster and how it should all work (in theory) and start in on how to set it all up (practically). In the end you'll have a light proxy setup that will rotate around between the servers but do it transparently from node to node.

PHP is designed using a shared-nothing architecture. [...] That's great! It makes for a very stable, very easy to use architecture. But what happens when you go beyond one server? [...] Managing a hundred servers is quite different from managing one. Heck, managing three servers is different from managing one.

Zend Server gives you all sorts of tools to help manage and monitor the server it runs on, but when you start clustering these servers, the maintenance can be a huge headache. The Cluster Manager comes in and helps with centralized configuration, monitoring, sessions clustering and more. Kevin briefly talks about setting it up (a call to yum install the package) and a few steps through a browser-based GUI interface to get the ball rolling. He's also included a screencast to show each step of the way.

The whitepaper talks about the changes in the PHP OCI8 1.3 extension, explains some of the concepts behind DRCP and FAN, and gives best practices and tuning tips. It includes a new PHP benchmark which shows up to 20,000 connections being handled by Oracle on commodity hardware using only 2G RAM.

The paper also talks about the FAN support that's built in - the ability for PHP to use the Oracle RAC cluster functionality to make for high availability (switching between nodes). The latest beta with all of this functionality in it can be grabbed from its page on the PECL site.