Microsoft Windows Server 2003 Deployment Kit [Electronic resources] : Planning Server Deployments نسخه متنی

اینجــــا یک کتابخانه دیجیتالی است

با بیش از 100000 منبع الکترونیکی رایگان به زبان فارسی ، عربی و انگلیسی

Microsoft Windows Server 2003 Deployment Kit [Electronic resources] : Planning Server Deployments - نسخه متنی

Microsoft Corporation

| نمايش فراداده ، افزودن یک نقد و بررسی
افزودن به کتابخانه شخصی
ارسال به دوستان
جستجو در متن کتاب
بیشتر
تنظیمات قلم

فونت

اندازه قلم

+ - پیش فرض

حالت نمایش

روز نیمروز شب
جستجو در لغت نامه
بیشتر
توضیحات
افزودن یادداشت جدید





Overview of the Server Clusters Design Process


Mission-critical applications, such as corporate databases and e-mail, must reside on systems that are designed for high availability and scalability. Deploying server clusters with the Microsoft Windows Server 2003, Enterprise Edition or the Windows Server 2003, Datacenter Edition operating system minimizes the amount of planned and unplanned server downtime. Server clusters can benefit your organization if:



Your users depend on regular access to mission-critical data and applications to do their jobs.



Your organization has established a limit on the amount of planned or unplanned service downtime that you can sustain.



The cost of the additional hardware that server clusters require is less than the cost of having mission-critical data and applications offline during a failure.



Windows Server 2003 provides two different clustering technologies: server clusters (discussed in this chapter) and Network Load Balancing. An overview of these cluster solutions is provided in "Planning for High Availability and Scalability" in this book. For more information about Network Load Balancing, see "Designing Network Load Balancing" and "Deploying Network Load Balancing" in this book. Note that this chapter focuses on using server clusters to make applications highly available. For information about using server clusters with resources such as file and print services, see "Designing and Deploying Print Servers" and "Designing and Deploying File Servers" in this book.

A third clustering technology, Component Load Balancing (CLB), is available in Microsoft Application Center 2000 but not in Windows Server 2003. CLB clusters enable COM+ applications to be distributed across multiple servers. For more information about CLB, see the documentation for Microsoft Application Center 2000.

If your business is considering server clusters to provide increased availability for one or more applications, and you plan to use Windows Server 2003 — either Enterprise Edition or Datacenter Edition — this chapter will help you evaluate, design, and implement a server cluster compatible with those applications. If you are already running server clusters on Windows NT 4.0 or Windows 2000, this chapter helps you prepare to upgrade those clusters to Windows Server 2003.



Server Cluster Design and Deployment Process


You begin the server cluster design and deployment process by defining your high-availability needs. After you determine the applications or services you want to host on a server cluster, you need to understand the clustering requirements of those applications or services. Next, design your server cluster support network, making sure that you protect your data from failure, disaster, or security risks. After you evaluate and account for all relevant hardware, software, network, and security factors, you are ready to deploy a server cluster. Figure 7.1 illustrates the server cluster design process.


Figure 7.1: Designing and Deploying Server Clusters



Server Cluster Fundamentals


The following sections provide an overview of key clustering concepts and summarize new clustering features that have been added to Windows Server 2003.

A cluster is a group of individual computer systems working together cooperatively to provide increased computing power and to ensure continuous availability of mission-critical applications or services. From the client viewpoint, an application that runs on a server cluster is no different than an application that runs on any other server, except that availability is higher.

Clustering Terms


The following terms are fundamental clustering concepts that are common to almost every clustering decision. For more information about the following terms and for an overview of server clusters, see "Server cluster components" and "Cluster objects" in Help and Support Center for Windows Server 2003.

Node A computer system that is a member of a server cluster. Windows Server 2003 supports up to eight nodes in a server cluster.

Resource A physical or logical entity that is capable of being managed by a cluster, brought online, taken offline, and moved between nodes. A resource can be owned only by a single node at any point in time.

Resource groups A collection of one or more resources that are managed and monitored as a single unit. Resource groups can be started and stopped independently of other groups (when a resource group is stopped, all resources within the group are stopped). In a server cluster, resource groups are indivisible units that are hosted on one node at any point in time. During failover, resource groups are transferred from one node to another.

Virtual server A collection of services that appear to clients as a physical Windows-based server but are not associated with a specific server. A virtual server is typically a resource group that contains all of the resources needed to run a particular application and can be failed over like any other resource group. All virtual servers must include a Network Name resource and an IP Address resource.

Failover The process of taking resource groups offline on one node and bringing them back online on another node. When a resource group goes offline, all resources belonging to that group go offline. The offline and online transitions occur in a predefined order. Resources that are dependent on other resources are taken offline before and brought online after the resources upon which they depend.


Fallback The process of moving resources, either individually or in a group, back to their original node after a failed node rejoins a cluster and comes back online.

Quorum resource The quorum-capable resource selected to maintain the configuration data necessary for recovery of the cluster. This data contains details of all of the changes that have been applied to the cluster database. The quorum resource is generally accessible to other cluster resources so that any cluster node has access to the most recent database changes. By default there is only one quorum resource per cluster.

Cluster Hardware Requirements


The following are general hardware requirements for building Windows Server 2003 clusters. More detailed information is provided throughout this chapter where applicable. In order for server clusters to be supported by Microsoft, all hardware must be selected from a list of qualified clustering solutions. See the Windows Server Catalog link on the Web Resources page at http://www.microsoft.com/windows/reskits/webresources. For more information about support for server clusters, see article Q309395, "The Microsoft Support Policy for Server Clusters and the Hardware." To find this article, see the Microsoft Knowledge Base link on the Web Resources page at http://www.microsoft.com/windows/reskits/webresources.



Use a minimum of two computers running Windows Server 2003, Enterprise Edition or Windows Server 2003, Datacenter Edition (computer hardware must also be listed in the Windows Server Catalog). You cannot mix x86-based and Itanium architecture-based computers in the same server cluster.



Meet minimum storage requirements. Minimum storage requirements depend on a number of factors, such as whether or not your server cluster uses storage area network (SAN) technology, or the type of quorum resource used in the server cluster. For more information about storage solutions, see "Designing the Support Structure for Server Clusters" later in this chapter. Note that all storage solutions must also be listed in the Windows Server Catalog.



Use a minimum of two network adapters for each node. In recommended configurations, one network adapter connects the node to the other nodes in the cluster for communication and configuration purposes (private network). The second adapter connects the cluster to both an external network (public network) and the private network.







Important

Server clusters do not protect against hardware failures. Microsoft recommends a redundant array of independent disks (RAID) solution for all cluster disks to eliminate disk drives as a potential single point of failure.



New in Windows Server 2003


Windows Server 2003 introduces a number of new clustering features that are highlighted below.

More nodes in a cluster Windows Server 2003 supports up to eight nodes per cluster.

Support for 64-bit versions of Windows Server 2003 operating system The 64-bit versions of Windows Server 2003, Enterprise Edition and Windows Server 2003, Datacenter Edition support the Cluster service. Note that you cannot use GUID partition table (GPT) disks for shared cluster storage. A GPT disk is an Itanium-based disk partition style in the 64-bit versions of Windows Server 2003. For 64-bit versions of Windows Server 2003, you must partition cluster disks on a shared bus as master boot record (MBR) disks and not as GPT disks.

Simplified cluster configuration and setup Server cluster system files are installed by default with Windows Server 2003. New features include the ability to create new server clusters — or add nodes to an existing cluster — remotely. Another new feature, the New Server Cluster Wizard, analyzes hardware and software configurations to identify potential problems before installation.

Security enhancements You can reset the Cluster service account password without stopping the Cluster service, allowing you to maintain corporate password policies without compromising availability. In addition, server clusters now support the Kerberos version 5 authentication protocol. For more information, see "Securing Server Clusters" later in this chapter.

Scripting An application can be made server cluster-aware through scripting (both VBScript and Jscript are supported), rather than through resource dynamic-link library (DLL) files. Unlike DLLs, scripting does not require knowledge of the C or C++ programming languages, which means scripts are easier for developers and administrators to create and implement. Scripts are also easier to customize for your applications.

Majority node set clusters In every cluster, the quorum resource maintains all configuration data necessary for the recovery of the cluster. In majority node set clusters, the quorum data is stored on each node, allowing for, among other things, geographically dispersed clusters. For more information about quorums, see "Designing the Support Structure for Server Clusters" later in this chapter. For more information about majority node set clusters, see "Protecting Data from Failure or Disaster" later in this chapter.

/ 122