5.1 Introduction to Exchange 2000 optimization

In the process of deploying Exchange 2000 in an enterprise, you will be confronted with the case of scale-up models that advocate server consolidation, or of scale-out models that reach out to many distributed sites and data centers. Each of these deployments requires a different type of hardware and software configuration, which is discussed in this chapter.

Some of the optimization will have to be done regardless of the client type and topology but according to the actual server configuration. For instance, the use of clustering and the need for fast a reliable failover make necessary certain modifications in the Exchange 2000 configuration that are not present by default. Most of the optimization to the product itself is done by modifying the Active Directory configuration-naming context objects and attributes that relate to the server, the Storage Groups, or the databases. Microsoft, very wisely, has moved from the technique of modifying system registries to the use of parameters located in the Active Directory, thereby enabling changes to be made centrally and taking advantage of the Active Directory replication to distribute the changes throughout the enterprise network.

There are optimization scenarios that apply depending on the type of access and the Exchange Organization topology. For instance, if you favor the use of Internet protocols for message transport (as opposed to X.400 connectors) and client access (as opposed to MAPI clients), you need to make sure that the communication layer between the Internet Information Service process and...