Failover clustering, network load balancing drive high availability

Windows shops with mission-critical servers, applications and servers can use Windows Server 2008's failover clustering and network load balancing features to ensure high availability.

Service provider takeaway: Service providers can help customers ensure high availability of their servers, applications and services using failover clustering and load balancing in Windows Server 2008.

Most of your customers know business productivity and revenues can be drastically affected if a mission-critical server, application or service fails. Indeed, one of the main objectives for IT departments everywhere is providing high availability for mission-critical resources. Toward that goal, service providers can implement high-availability alternatives in Windows Server 2008 to mitigate server outages for their Windows shop customers.

The first step in designing a Windows-based high-availability solution entails understanding the two main high-availability alternatives available with Windows Server 2008; failover clustering and network load balancing. These options tackle high availability in different ways.

Failover clustering

At the macro level, a Windows Server 2008 failover cluster provides high availability by eliminating the threat of a single point of failure for a server, application or service. Normally, if a server with a particular application or service crashes, the application or service is unavailable until an administrator manually rectifies the problem. But if a clustered server crashes, another server within the cluster will automatically take over the failed server's application and service responsibilities without intervention from an administrator or impact on operations.

Windows Server 2008 supports the shared-nothing cluster model, in which two or more independent servers, or nodes, share resources; each server owns and is responsible for managing its local resources and provides nonsharing services. In case of a node failure, the disks, resources and services running on the failed node fail over to a surviving node in the cluster. For example, if an Exchange server is operating on node 1 of the cluster and it crashes, the Exchange application and services will automatically fail over to node 2 of the cluster. This model minimizes server outage and downtime. Only one node manages one particular set of disks, cluster resources and services at any given time.

When designing and implementing failover clusters, service providers need to ensure the following preconditions: that each server's hardware specifications are identical, that a shared storage server such as a SAN or NAS is in place, and that a dedicated network for heartbeat communication between server nodes is available. In addition, all hardware and software drivers associated with the cluster must be certified by Microsoft, and the customer must use either the Enterprise or Data Center Edition of Windows Server 2008. Those editions support as many as 16 nodes in a single failover cluster implementation.

Network load balancing
Network load balancing (NLB), Windows Server 2008's other high-availability alternative, enables an organization to scale server and application performance by distributing TCP/IP requests to multiple servers, also known as hosts, within a server farm. This scenario optimizes resource utilization, decreases computing time and ensures server availability. Typically, service providers should consider network load balancing if their customer situation includes, but is not limited to, Web server farms, Terminal Services farms, media servers or Exchange Outlook Web Access servers.

Above and beyond providing scalability by distributing TCP/IP traffic among servers participating in a farm, NLB also ensures high availability by identifying host failures and automatically redistributing traffic to the surviving hosts.

Network load balancing is native to all editions of Windows Server 2008. Unlike failover clustering, NLB does not require any special hardware, and a network load balancing server farm can include as many as 32 nodes. When designing and implementing NLB server farms, it's common to start off with two servers for scalability and high availability and then add additional nodes to the farm as TCP/IP traffic increases.

More on high availability
High-availability systems need the right tools, skills

Three low-cost approaches to high-availability databases

Clearly, failover clustering and network load balancing with Windows Server 2008 provide service providers with options when designing and implementing high availability for their customers' mission-critical servers and applications. Through the use of failover clustering and network load balancing, customers will gain an increase in server availability to mission-critical servers, a decrease in downtime during routine maintenance, a decrease in server outages, and a minimization of end-user outages during a failover.

About the author
Ross Mistry is a partner and principal consultant at Convergent Computing, located in the San Francisco Bay area. He is a co-author of SQL Server 2005 Management & Administration and Windows Server 2008 Unleashed. Ross frequently speaks at international conferences such as SQL Server PASS and Dev Connections. He is currently working on his latest title, SQL Server 2008 Management & Administration -- Mastering the Database Engine, which is scheduled for release in Fall 2008.


Dig Deeper on Operating Systems and Software Services