Chapter 6 – Exchange Server 2007 Failover Clustering

Publisher Summary This chapter provides an introduction to Windows Failover Clustering (WFC) in Windows Server 2003 and Windows Server 2008, followed by a discussion of Exchange Server 2007 Clustered Mailbox Servers. Clusters are multiple computers that appear to users as a single highly available system. The objective of clustering concept is a “no single point of failure” configuration. If one component fails, another component takes over the responsibility, and the service offered by the cluster should not be affected from an end-user point of view. Typically, a clustered deployment is considered if one wants to increase the availability of messaging service, or if business continuity requirements demand a geographically dispersed deployment with relatively short recovery time objectives. Exchange administrators should have at least a basic knowledge of the WFC concept to understand the various cluster architectures used by the two implementation alternatives for a Clustered Mailbox Server (CMS) in Exchange Server 2007, Cluster Continuous Replication (CCR), and Single Copy Cluster (SCC). The mailbox server role is the only role in Exchange Server 2007 that can be deployed on a cluster. In Exchange Server 2007, Outlook Web Access or transport services are provided by the Client Access Server (CAS) and Hub Transport (HT) server role. Both roles cannot be installed on a WFC. A server with the HT and another with the CAS role must be online in the same AD site; otherwise, users cannot send and receive emails or connect to their mailbox using Internet protocols. The minimal configuration is four Exchange servers: two clustered nodes for the mailbox server and two multirole servers with both the HT and CAS role installed. Additionally, two Active Directory (AD) global catalog servers are needed.