Chapter 5. Planning, Implementing, and Maintaining Highly Available Servers

O BJECTIVES

High availability has become something of a buzzword in the past few years in the IT sector. People were striving for many years before that to achieve "Five Nines," or 99.999% server uptime. Microsoft makes it easier than ever before to achieve this goal with the Network Load Balancing and Clustering services provided in Windows Server 2003.

Microsoft has the following objectives for the high availability portion of the "Planning, Implementing, and Maintaining Server Availability" unit:

Plan services for high availability.

• Plan a high availability solution that uses clustering services.

• Plan a high availability solution that uses network load balancing.

• High availablity solutions are becoming more popular and more critical in today's Internet-based economy. It is important to understand and be able to work with the highly available solutions provided with Windows Server 2003.

Implement a cluster server.

• After finishing your plan for a high availability cluster solution, you need to successfully implement and manage it.

Monitor network load balancing. Tools might include the Network Load Balancing Monitor Microsoft Management Console ( MMC ) snap-in and the WLBS cluster control utility.

• Network load balancing (NLB) typically can be counted on to function with little to no administrative action required. You do, however, need to perform routine monitoring on your NLB clusters to ensure that they are operating properly. Windows Server 2003 contains several useful utilities that you can use to monitor NLB.

Recover from cluster node failure.

• Clusters are implemented to minimize the adverse effects of failure. The capability to quickly and effectively recover from the failure of one or more cluster nodes is a key part of the overall high availability solution.

O UTLINE

Introduction

Designing for High Availability

Uninterruptible Power Supplies

Fault-Tolerant Storage Systems

Redundant Networking Infrastructure

Planning High Availability Solutions

High Availability Solutions

High Availability Terminology

Planning and Implementing NLB Clusters

Port Rules

Filtering Mode and Affinity

NLB Cluster Operation Mode

Creating an NLB Cluster

Planning and Implementing MSCS Clusters

Cluster Models

Single Node Cluster

Single Quorum Cluster

Majority Node Set Cluster

Cluster Operation Modes

Cluster Failover Polices

Creating a Cluster

Monitoring and Managing High Availability Solutions

Recovering from Failed Cluster Nodes

Monitoring Network Load Balancing

Chapter Summary

Apply Your Knowledge

Exercises

Review Questions

Exam Questions

Answers to Review Questions

Answers to Exam Questions

Suggested Readings and Resources

S TUDY S TRATEGIES

• For many people, the concept of high availability servers is a new one. It is important to get a good grasp of the concepts presented in this chapter, including high availability, clustering, network load balancing, failover, and failback.

• Ensure that you understand the specific requirements, benefits, and costs associated with high availability solutions utilizing clustering or network load balancing.

• Get your hands dirty. The Step by Steps throughout this book provide plenty of directions and exercises, but you should go beyond these examples and create some of your own. If you can, experiment with each of the objectives to see how they work and why you would use each one.