When setting up a Microsoft Exchange infrastructure you need to consider the network topology (or network layout). This includes servers, switches, firewalls, and hardware LoadBalancers.

This being a typical beginner question we will explain a few simple examples, discuss potential difficulties and advanced setups. High Availability is a must in this context. Therefore the examples will include two LoadBalancers (“High Availability pair” or “HA-cluster”).

May we introduce our newest product: The GEO LoadMaster (GLM). Instead of a local load balancing server we have developed a server for access control to several geographical locations.

Here you can see a few examples for the use of a “Global Server Load Balancer”:

It’s a few years old, but I have only just found it: a really nice introduction to this subject is offer by both these articles (part 1 and part 2). Author Gregor Roth differentiates between “Transport-Level” and “Application-Level” and puts his illustration in a context of Java-servers and Java-programming.  Apart from that the explanations are globally valid.

After some really technical articles, I would like to take a step back and talk about some fundamental questions:  How can I use Load Balancers for my (web based or other) applications? Which are the criteria for a good product?

Actually these are not quite the right questions. In reality we need to solve an existing problem and not find a problem for an existing tool. So it’s more like this:

SSL-Acceleration is one of the biggest advantages of a suitable Load Balancer – for many users actually the principal purpose. But does it work in any case? Yes it does, if some requirements are fulfilled. First let’s have a look at what SSL-Acceleration means.