|
|
 |
|
|
|
Microsoft® SQL Server 2000™ Performance Tuning Technical Reference
|
|
|
Author
|
|
Edward Whalen, Marcilina Garcia, Steve Adrien DeLuca, Dean Thompson
|
|
|
Pages
|
464
|
|
Disk
|
N/A
|
|
Level
|
Int/Adv
|
|
Published
|
07/03/2001
|
|
ISBN
|
9780735612709
|
|
Price
|
$49.99
To see this book's discounted price, select a reseller below.
|
|
|
|
|
|
|
|
|
Chapter 5: New Features and Performance Enhancements continued
SQL Server 2000 Failover Clustering
SQL Server 2000 is a cluster-aware application that can be configured to provide failover capabilities. You must first install Microsoft Cluster Server (MSCS), which comes with the Windows 2000 Advanced Server and Datacenter Server operating systems only. Also, you must have SQL Server 2000 Enterprise Edition to support failover clustering. With failover clustering, if one system fails, SQL Server fails, or the system is taken offline, SQL Server processing switches to a second, clustered system—this switch is known as a failover. The goal of failover is to minimize system downtime. After a failover, you can restore the failed system and resume processing by switching back to the restored system—this is called failback.
Each system that is part of a cluster is called a node. Cluster Server with Windows 2000 Advanced Server supports up to two nodes in a cluster, and with Windows 2000 Datacenter Server it supports up to four nodes in a cluster. The nodes in a cluster share a common set of cluster resources, such as disk drives. Each node is connected to a common network and can communicate with all other nodes. The nodes send each other network messages called heartbeat messages. If the MSCS software detects the loss of a heartbeat from one of the nodes in the cluster, failover occurs for that node.
SQL Server 7.0 also supported failover clustering, but the administration of failover clustering is greatly improved in SQL Server 2000. SQL Server failover setup is no longer performed by running the Failover Cluster wizard, instead it is part of the SQL Server Setup program. First you must have installed the Microsoft Cluster Server service and configured it with the Cluster Service Configuration wizard on Windows 2000 Advanced Server or Datacenter Server. Before you start the SQL Server installation, create the same path on each node for the location where you want to install SQL Server program files. Then you can install SQL Server with the Setup program. At the start of the SQL Server installation, the Setup program asks several questions regarding the cluster configuration. Then the regular SQL Server installation continues. After you’ve entered all the information, the Setup program copies SQL Server binaries (SQL program files) to all of the server nodes in the cluster, onto the local drive of each node in the path you specified during installation. This is why it’s important to create the same path on each node before you start the installation. With this design, when a failover occurs with SQL Server 2000, only the databases are failed over, not the SQL Server binaries. (In SQL Server 7.0 the binaries were failed over also.)
The new administration tasks you can perform with failover clustering include:
- Administer failover clustering from any node in the cluster through Cluster Administrator
- Allow one cluster node to fail over to any other node in the cluster (assuming you have a three or four-node cluster with Windows 2000 Datacenter Server)
- Reinstall or rebuild a cluster instance on any node in the cluster without affecting the other cluster node instances
- Specify multiple IP addresses for a virtual server
- Add or remove nodes from the failover cluster by using SQL Server Setup
- Fail over or fail back to or from any node in the cluster
For more information on using Microsoft Cluster Server and steps for installing SQL Server for failover, see SQL Server Books Online.
Previous
| Table of Contents
|
Next
Last Updated: Saturday, July 7, 2001 |
