MOC Course 6423A Implementing & Managing Windows Server 2008 Clustering

O F F I C I A L M I C R O S O F T L E A R N I N G P R O D U C T

6423A Implementing and Managing Windows Server®°2008 Clustering

ii Implementing and Managing Windows Server® 2008 Clustering

Information in this document, including URL and other Internet Web site references, is subject to change without notice. Unless otherwise noted, the example companies, organizations, products, domain names, e-mail addresses, logos, people, places, and events depicted herein are fictitious, and no association with any real company, organization, product, domain name, e-mail address, logo, person, place or event is intended or should be inferred. Complying with all applicable copyright laws is the responsibility of the user. Without limiting the rights under copyright, no part of this document may be reproduced, stored in or introduced into a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording, or otherwise), or for any purpose, without the express written permission of Microsoft Corporation.

Microsoft may have patents, patent applications, trademarks, copyrights, or other intellectual property rights covering subject matter in this document. Except as expressly provided in any written license agreement from Microsoft, the furnishing of this document does not give you any license to these patents, trademarks, copyrights, or other intellectual property.

The names of manufacturers, products, or URLs are provided for informational purposes only and Microsoft makes no representations and warranties, either expressed, implied, or statutory, regarding these manufacturers or the use of the products with any Microsoft technologies. The inclusion of a manufacturer or product does not imply endorsement of Microsoft of the manufacturer or product. Links may be provided to third party sites. Such sites are not under the control of Microsoft and Microsoft is not responsible for the contents of any linked site or any link contained in a linked site, or any changes or updates to such sites. Microsoft is not responsible for webcasting or any other form of transmission received from any linked site. Microsoft is providing these links to you only as a convenience, and the inclusion of any link does not imply endorsement of Microsoft of the site or the products contained therein.

© 2008 Microsoft Corporation. All rights reserved.

Microsoft, Active Directory, Hyper-V, Internet Explorer, Outlook, PowerPoint, SQL Server, Windows, and Windows Vista are either registered trademarks or trademarks of Microsoft Corporation in the United States and/or other countries.

All other trademarks are property of their respective owners.

Technical Reviewer: Rogerio Melo

Product Number: 6423A

Part Number: X14-89109

Released: 05/2008

MICROSOFT LICENSE TERMS OFFICIAL MICROSOFT LEARNING PRODUCTS - TRAINER EDITION – Pre-Release and Final Release Versions These license terms are an agreement between Microsoft Corporation and you. Please read them. They apply to the Licensed Content named above, which includes the media on which you received it, if any. The terms also apply to any Microsoft

• updates,

• supplements,

• Internet-based services, and

• support services

for this Licensed Content, unless other terms accompany those items. If so, those terms apply.

By using the Licensed Content, you accept these terms. If you do not accept them, do not use the Licensed Content.

If you comply with these license terms, you have the rights below.

1. DEFINITIONS.

a. “Academic Materials” means the printed or electronic documentation such as manuals, workbooks, white papers, press releases, datasheets, and FAQs which may be included in the Licensed Content.

b. “Authorized Learning Center(s)” means a Microsoft Certified Partner for Learning Solutions location, an IT Academy location, or such other entity as Microsoft may designate from time to time.

c. “Authorized Training Session(s)” means those training sessions authorized by Microsoft and conducted at or through Authorized Learning Centers by a Trainer providing training to Students solely on Official Microsoft Learning Products (formerly known as Microsoft Official Curriculum or “MOC”) and Microsoft Dynamics Learning Products (formerly know as Microsoft Business Solutions Courseware). Each Authorized Training Session will provide training on the subject matter of one (1) Course.

d. “Course” means one of the courses using Licensed Content offered by an Authorized Learning Center during an Authorized Training Session, each of which provides training on a particular Microsoft technology subject matter.

e. “Device(s)” means a single computer, device, workstation, terminal, or other digital electronic or analog device.

f. “Licensed Content” means the materials accompanying these license terms. The Licensed Content may include, but is not limited to, the following elements: (i) Trainer Content, (ii) Student Content, (iii) classroom setup guide, and (iv) Software. There are different and separate components of the Licensed Content for each Course.

g. “Software” means the Virtual Machines and Virtual Hard Disks, or other software applications that may be included with the Licensed Content.

h. “Student(s)” means a student duly enrolled for an Authorized Training Session at your location.

i. “Student Content” means the learning materials accompanying these license terms that are for use by Students and Trainers during an Authorized Training Session. Student Content may include labs, simulations, and courseware files for a Course.

j. “Trainer(s)” means a) a person who is duly certified by Microsoft as a Microsoft Certified Trainer and b) such other individual as authorized in writing by Microsoft and has been engaged by an Authorized Learning Center to teach or instruct an Authorized Training Session to Students on its behalf.

k. “Trainer Content” means the materials accompanying these license terms that are for use by Trainers and Students, as applicable, solely during an Authorized Training Session. Trainer Content may include Virtual Machines, Virtual Hard Disks, Microsoft PowerPoint files, instructor notes, and demonstration guides and script files for a Course.

l. “Virtual Hard Disks” means Microsoft Software that is comprised of virtualized hard disks (such as a base virtual hard disk or differencing disks) for a Virtual Machine that can be loaded onto a single computer or other device in order to allow end-users to run multiple operating systems concurrently. For the purposes of these license terms, Virtual Hard Disks will be considered “Trainer Content”.

m. “Virtual Machine” means a virtualized computing experience, created and accessed using Microsoft® Virtual PC or Microsoft® Virtual Server software that consists of a virtualized hardware environment, one or more Virtual Hard Disks, and a configuration file setting the parameters of the virtualized hardware environment (e.g., RAM). For the purposes of these license terms, Virtual Hard Disks will be considered “Trainer Content”.

n. “you” means the Authorized Learning Center or Trainer, as applicable, that has agreed to these license terms.

2. OVERVIEW.

Licensed Content. The Licensed Content includes Software, Academic Materials (online and electronic), Trainer Content, Student Content, classroom setup guide, and associated media.

License Model. The Licensed Content is licensed on a per copy per Authorized Learning Center location or per Trainer basis.

3. INSTALLATION AND USE RIGHTS.

a. Authorized Learning Centers and Trainers: For each Authorized Training Session, you may:

i. either install individual copies of the relevant Licensed Content on classroom Devices only for use by Students enrolled in and the Trainer delivering the Authorized Training Session, provided that the number of copies in use does not exceed the number of Students enrolled in and the Trainer delivering the Authorized Training Session, OR

ii. install one copy of the relevant Licensed Content on a network server only for access by classroom Devices and only for use by Students enrolled in and the Trainer delivering the Authorized Training Session, provided that the number of Devices accessing the Licensed Content on such server does not exceed the number of Students enrolled in and the Trainer delivering the Authorized Training Session.

iii. and allow the Students enrolled in and the Trainer delivering the Authorized Training Session to use the Licensed Content that you install in accordance with (ii) or (ii) above during such Authorized Training Session in accordance with these license terms.

i. Separation of Components. The components of the Licensed Content are licensed as a single unit. You may not separate the components and install them on different Devices.

ii. Third Party Programs. The Licensed Content may contain third party programs. These license terms will apply to the use of those third party programs, unless other terms accompany those programs.

b. Trainers:

i. Trainers may Use the Licensed Content that you install or that is installed by an Authorized Learning Center on a classroom Device to deliver an Authorized Training Session.

ii. Trainers may also Use a copy of the Licensed Content as follows:

A. Licensed Device. The licensed Device is the Device on which you Use the Licensed Content. You may install and Use one copy of the Licensed Content on the licensed Device solely for your own personal training Use and for preparation of an Authorized Training Session.

B. Portable Device. You may install another copy on a portable device solely for your own personal training Use and for preparation of an Authorized Training Session.

4. PRE-RELEASE VERSIONS. If this is a pre-release (“beta”) version, in addition to the other provisions in this agreement, these terms also apply:

a. Pre-Release Licensed Content. This Licensed Content is a pre-release version. It may not contain the same information and/or work the way a final version of the Licensed Content will. We may change it for the final, commercial version. We also may not release a commercial version. You will clearly and conspicuously inform any Students who participate in each Authorized Training Session of the foregoing; and, that you or Microsoft are under no obligation to provide them with any further content, including but not limited to the final released version of the Licensed Content for the Course.

b. Feedback. If you agree to give feedback about the Licensed Content to Microsoft, you give to Microsoft, without charge, the right to use, share and commercialize your feedback in any way and for any purpose. You also give to third parties, without charge, any patent rights needed for their products, technologies and services to use or interface with any specific parts of a Microsoft software, Licensed Content, or service that includes the feedback. You will not give feedback that is subject to a license that requires Microsoft to license its software or documentation to third parties because we include your feedback in them. These rights survive this agreement.

c. Confidential Information. The Licensed Content, including any viewer, user interface, features and documentation that may be included with the Licensed Content, is confidential and proprietary to Microsoft and its suppliers.

i. Use. For five years after installation of the Licensed Content or its commercial release, whichever is first, you may not disclose confidential information to third parties. You may disclose confidential information only to your employees and consultants who need to know the information. You must have written agreements with them that protect the confidential information at least as much as this agreement.

ii. Survival. Your duty to protect confidential information survives this agreement.

iii. Exclusions. You may disclose confidential information in response to a judicial or governmental order. You must first give written notice to Microsoft to allow it to seek a

protective order or otherwise protect the information. Confidential information does not include information that

• becomes publicly known through no wrongful act;

• you received from a third party who did not breach confidentiality obligations to Microsoft or its suppliers; or

• you developed independently.

d. Term. The term of this agreement for pre-release versions is (i) the date which Microsoft informs you is the end date for using the beta version, or (ii) the commercial release of the final release version of the Licensed Content, whichever is first (“beta term”).

e. Use. You will cease using all copies of the beta version upon expiration or termination of the beta term, and will destroy all copies of same in the possession or under your control and/or in the possession or under the control of any Trainers who have received copies of the pre-released version.

f. Copies. Microsoft will inform Authorized Learning Centers if they may make copies of the beta version (in either print and/or CD version) and distribute such copies to Students and/or Trainers. If Microsoft allows such distribution, you will follow any additional terms that Microsoft provides to you for such copies and distribution.

5. ADDITIONAL LICENSING REQUIREMENTS AND/OR USE RIGHTS.

a. Authorized Learning Centers and Trainers:

i. Software.

ii. Virtual Hard Disks. The Licensed Content may contain versions of Microsoft XP, Microsoft Windows Vista, Windows Server 2003, Windows Server 2008, and Windows 2000 Advanced Server and/or other Microsoft products which are provided in Virtual Hard Disks.

A. If the Virtual Hard Disks and the labs are launched through the Microsoft Learning Lab Launcher, then these terms apply:

Time-Sensitive Software. If the Software is not reset, it will stop running based upon the time indicated on the install of the Virtual Machines (between 30 and 500 days after you install it). You will not receive notice before it stops running. You may not be able to access data used or information saved with the Virtual Machines when it stops running and may be forced to reset these Virtual Machines to their original state. You must remove the Software from the Devices at the end of each Authorized Training Session and reinstall and launch it prior to the beginning of the next Authorized Training Session.

B. If the Virtual Hard Disks require a product key to launch, then these terms apply:

Microsoft will deactivate the operating system associated with each Virtual Hard Disk. Before installing any Virtual Hard Disks on classroom Devices for use during an Authorized Training Session, you will obtain from Microsoft a product key for the operating system software for the Virtual Hard Disks and will activate such Software with Microsoft using such product key.

C. These terms apply to all Virtual Machines and Virtual Hard Disks:

You may only use the Virtual Machines and Virtual Hard Disks if you comply with the terms and conditions of this agreement and the following security requirements:

o You may not install Virtual Machines and Virtual Hard Disks on portable Devices or Devices that are accessible to other networks.

o You must remove Virtual Machines and Virtual Hard Disks from all classroom Devices at the end of each Authorized Training Session, except those held at Microsoft Certified Partners for Learning Solutions locations.

o You must remove the differencing drive portions of the Virtual Hard Disks from all classroom Devices at the end of each Authorized Training Session at Microsoft Certified Partners for Learning Solutions locations.

o You will ensure that the Virtual Machines and Virtual Hard Disks are not copied or downloaded from Devices on which you installed them.

o You will strictly comply with all Microsoft instructions relating to installation, use, activation and deactivation, and security of Virtual Machines and Virtual Hard Disks.

o You may not modify the Virtual Machines and Virtual Hard Disks or any contents thereof.

o You may not reproduce or redistribute the Virtual Machines or Virtual Hard Disks.

ii. Classroom Setup Guide. You will assure any Licensed Content installed for use during an Authorized Training Session will be done in accordance with the classroom set-up guide for the Course.

iii. Media Elements and Templates. You may allow Trainers and Students to use images, clip art, animations, sounds, music, shapes, video clips and templates provided with the Licensed Content solely in an Authorized Training Session. If Trainers have their own copy of the Licensed Content, they may use Media Elements for their personal training use.

iv. iv Evaluation Software. Any Software that is included in the Student Content designated as “Evaluation Software” may be used by Students solely for their personal training outside of the Authorized Training Session.

b. Trainers Only:

i. Use of PowerPoint Slide Deck Templates. The Trainer Content may include Microsoft PowerPoint slide decks. Trainers may use, copy and modify the PowerPoint slide decks only for providing an Authorized Training Session. If you elect to exercise the foregoing, you will agree or ensure Trainer agrees: (a) that modification of the slide decks will not constitute creation of obscene or scandalous works, as defined by federal law at the time the work is created; and (b) to comply with all other terms and conditions of this agreement.

ii. Use of Instructional Components in Trainer Content. For each Authorized Training Session, Trainers may customize and reproduce, in accordance with the MCT Agreement, those portions of the Licensed Content that are logically associated with instruction of the Authorized Training Session. If you elect to exercise the foregoing rights, you agree or ensure the Trainer agrees: (a) that any of these customizations or reproductions will only be used for providing an Authorized Training Session and (b) to comply with all other terms and conditions of this agreement.

iii. Academic Materials. If the Licensed Content contains Academic Materials, you may copy and use the Academic Materials. You may not make any modifications to the Academic Materials and you may not print any book (either electronic or print version) in its entirety. If you reproduce any Academic Materials, you agree that:

• The use of the Academic Materials will be only for your personal reference or training use

• You will not republish or post the Academic Materials on any network computer or broadcast in any media;

• You will include the Academic Material’s original copyright notice, or a copyright notice to Microsoft’s benefit in the format provided below:

Form of Notice:

© 2008 Reprinted for personal reference use only with permission by Microsoft Corporation. All rights reserved.

Microsoft, Windows, and Windows Server are either registered trademarks or trademarks of Microsoft Corporation in the US and/or other countries. Other product and company names mentioned herein may be the trademarks of their respective owners.

6. INTERNET-BASED SERVICES. Microsoft may provide Internet-based services with the Licensed Content. It may change or cancel them at any time. You may not use these services in any way that could harm them or impair anyone else’s use of them. You may not use the services to try to gain unauthorized access to any service, data, account or network by any means.

7. SCOPE OF LICENSE. The Licensed Content is licensed, not sold. This agreement only gives you some rights to use the Licensed Content. Microsoft reserves all other rights. Unless applicable law gives you more rights despite this limitation, you may use the Licensed Content only as expressly permitted in this agreement. In doing so, you must comply with any technical limitations in the Licensed Content that only allow you to use it in certain ways. You may not

• install more copies of the Licensed Content on classroom Devices than the number of Students and the Trainer in the Authorized Training Session;

• allow more classroom Devices to access the server than the number of Students enrolled in and the Trainer delivering the Authorized Training Session if the Licensed Content is installed on a network server;

• copy or reproduce the Licensed Content to any server or location for further reproduction or distribution;

• disclose the results of any benchmark tests of the Licensed Content to any third party without Microsoft’s prior written approval;

• work around any technical limitations in the Licensed Content;

• reverse engineer, decompile or disassemble the Licensed Content, except and only to the extent that applicable law expressly permits, despite this limitation;

• make more copies of the Licensed Content than specified in this agreement or allowed by applicable law, despite this limitation;

• publish the Licensed Content for others to copy;

• transfer the Licensed Content, in whole or in part, to a third party;

• access or use any Licensed Content for which you (i) are not providing a Course and/or (ii) have not been authorized by Microsoft to access and use;

• rent, lease or lend the Licensed Content; or

• use the Licensed Content for commercial hosting services or general business purposes.

• Rights to access the server software that may be included with the Licensed Content, including the Virtual Hard Disks does not give you any right to implement Microsoft patents or other Microsoft intellectual property in software or devices that may access the server.

8. EXPORT RESTRICTIONS. The Licensed Content is subject to United States export laws and regulations. You must comply with all domestic and international export laws and regulations that apply to the Licensed Content. These laws include restrictions on destinations, end users and end use. For additional information, see www.microsoft.com/exporting.

9. NOT FOR RESALE SOFTWARE/LICENSED CONTENT. You may not sell software or Licensed Content marked as “NFR” or “Not for Resale.”

10. ACADEMIC EDITION. You must be a “Qualified Educational User” to use Licensed Content marked as “Academic Edition” or “AE.” If you do not know whether you are a Qualified Educational User, visit www.microsoft.com/education or contact the Microsoft affiliate serving your country.

11. TERMINATION. Without prejudice to any other rights, Microsoft may terminate this agreement if you fail to comply with the terms and conditions of these license terms. In the event your status as an Authorized Learning Center or Trainer a) expires, b) is voluntarily terminated by you, and/or c) is terminated by Microsoft, this agreement shall automatically terminate. Upon any termination of this agreement, you must destroy all copies of the Licensed Content and all of its component parts.

12. ENTIRE AGREEMENT. This agreement, and the terms for supplements, updates, Internet-based services and support services that you use, are the entire agreement for the Licensed Content and support services.

13. APPLICABLE LAW.

a. United States. If you acquired the Licensed Content in the United States, Washington state law governs the interpretation of this agreement and applies to claims for breach of it, regardless of conflict of laws principles. The laws of the state where you live govern all other claims, including claims under state consumer protection laws, unfair competition laws, and in tort.

b. Outside the United States. If you acquired the Licensed Content in any other country, the laws of that country apply.

14. LEGAL EFFECT. This agreement describes certain legal rights. You may have other rights under the laws of your country. You may also have rights with respect to the party from whom you acquired the Licensed Content. This agreement does not change your rights under the laws of your country if the laws of your country do not permit it to do so.

15. DISCLAIMER OF WARRANTY. The Licensed Content is licensed “as-is.” You bear the risk of using it. Microsoft gives no express warranties, guarantees or conditions. You may have additional consumer rights under your local laws which this agreement cannot change. To the extent permitted under your local laws, Microsoft excludes the implied warranties of merchantability, fitness for a particular purpose and non-infringement.

16. LIMITATION ON AND EXCLUSION OF REMEDIES AND DAMAGES. YOU CAN RECOVER FROM MICROSOFT AND ITS SUPPLIERS ONLY DIRECT DAMAGES UP TO U.S. $5.00. YOU CANNOT RECOVER ANY OTHER DAMAGES, INCLUDING CONSEQUENTIAL, LOST PROFITS, SPECIAL, INDIRECT OR INCIDENTAL DAMAGES.

This limitation applies to

• anything related to the Licensed Content, software, services, content (including code) on third party Internet sites, or third party programs; and

• claims for breach of contract, breach of warranty, guarantee or condition, strict liability, negligence, or other tort to the extent permitted by applicable law.

It also applies even if Microsoft knew or should have known about the possibility of the damages. The above limitation or exclusion may not apply to you because your country may not allow the exclusion or limitation of incidental, consequential or other damages.

Please note: As this Licensed Content is distributed in Quebec, Canada, some of the clauses in this agreement are provided below in French.

Remarque : Ce le contenu sous licence étant distribué au Québec, Canada, certaines des clauses dans ce contrat sont fournies ci-dessous en français.

EXONÉRATION DE GARANTIE. Le contenu sous licence visé par une licence est offert « tel quel ». Toute utilisation de ce contenu sous licence est à votre seule risque et péril. Microsoft n’accorde aucune autre garantie expresse. Vous pouvez bénéficier de droits additionnels en vertu du droit local sur la protection dues consommateurs, que ce contrat ne peut modifier. La ou elles sont permises par le droit locale, les garanties implicites de qualité marchande, d’adéquation à un usage particulier et d’absence de contrefaçon sont exclues.

LIMITATION DES DOMMAGES-INTÉRÊTS ET EXCLUSION DE RESPONSABILITÉ POUR LES DOMMAGES. Vous pouvez obtenir de Microsoft et de ses fournisseurs une indemnisation en cas de dommages directs uniquement à hauteur de 5,00 $ US. Vous ne pouvez prétendre à aucune indemnisation pour les autres dommages, y compris les dommages spéciaux, indirects ou accessoires et pertes de bénéfices.

Cette limitation concerne:

• tout ce qui est relié au le contenu sous licence , aux services ou au contenu (y compris le code) figurant sur des sites Internet tiers ou dans des programmes tiers ; et

• les réclamations au titre de violation de contrat ou de garantie, ou au titre de responsabilité stricte, de négligence ou d’une autre faute dans la limite autorisée par la loi en vigueur.

Elle s’applique également, même si Microsoft connaissait ou devrait connaître l’éventualité d’un tel dommage. Si votre pays n’autorise pas l’exclusion ou la limitation de responsabilité pour les dommages indirects, accessoires ou de quelque nature que ce soit, il se peut que la limitation ou l’exclusion ci-dessus ne s’appliquera pas à votre égard.

EFFET JURIDIQUE. Le présent contrat décrit certains droits juridiques. Vous pourriez avoir d’autres droits prévus par les lois de votre pays. Le présent contrat ne modifie pas les droits que vous confèrent les lois de votre pays si celles-ci ne le permettent pas.

Implementing and Managing Windows Server® 2008 Clustering xi

Contents

Module 1: Introduction to Clusters

Lesson 1: Overview of Clusters 1-3

Lesson 2: Benefits of Using Clusters 1-11

Lesson 3: Overview of the Windows Server 2008 High Availability

Solutions 1-20

Lab: Identifying Windows Server 2008 High-Availability Solutions 1-29

Module 2: Introduction to Microsoft Windows Server 2008 Failover Clusters

Lesson 1: Overview of Windows Server 2008 Failover Clusters 2-3

Lesson 2: Key Windows Server 2008 Failover Cluster Features 2-12

Lesson 3: Overview of the Windows Server 2008 Quorum Modes 2-23

Lab: Identifying Windows Server 2008 Clustering Solutions 2-36

Module 3: Preparing to Install a Failover Cluster

Lesson 1: Overview of Requirements for Installing a Failover Cluster 3-3

Lesson 2: Planning the Failover Cluster Implementation 3-13

Lesson 3: Installing the Failover Clustering Feature and Validating the

Cluster Configuration 3-26

Lesson 4: Installing the Failover Cluster on Windows Server 2008

Server Core 3-41

Lab: Preparing for a Cluster Installation 3-47

Module 4: Overview of Failover Cluster Storage Requirements

Lesson 1: Overview of Storage Technologies 4-3

Lesson 2: Introduction to Storage Area Networks 4-22

Lesson 3: Planning a Storage Solution for Failover Clusters 4-37

Lesson 4: Configuring an iSCSI Storage Connection 4-43

Lab: Identifying SAN Components 4-49

xii Implementing and Managing Windows Server® 2008 Clustering

Module 5: Configuring a Failover Cluster

Lesson 1: Creating a New Failover Cluster 5-3

Lesson 2: Managing a Failover Cluster 5-7

Lesson 3: Verifying Failover Functionality 5-14

Lab: Creating and Administering a Cluster 5-21

Module 6: Configuring Cluster Resources and Server Roles

Lesson 1: Configuring Cluster Resources 6-3

Lesson 2: Implementing Failover Clusters for Server Roles Using

Failover Cluster Management 6-9

Lesson 3: Clustering Server Roles Using Windows Server Core 6-15

Lab: Clustering Server Roles and Features 6-22

Module 7: Maintaining Microsoft Failover Clusters

Lesson 1: Monitoring Failover Clusters 7-3

Lesson 2: Backing Up and Restoring Failover Clusters 7-11

Lesson 3: Troubleshooting Failover Clusters 7-20

Lab: Maintaining Failover Clusters 7-26

Module 8: Implementing Multi-Site Clusters

Lesson 1: Overview of Multi-Site Clusters 8-3

Lesson 2: Implementing Multi-Site Clusters Using Windows Server 2008 8-14

Module 9: Implementing Network Load Balancing Clusters

Lesson 1: Overview of Network Load Balancing Clusters 9-3

Lesson 2: Configuring a Network Load Balancing Cluster 9-15

Lesson 3: Maintaining a Network Load Balancing Cluster 9-30

Lab: Implementing an NLB Cluster 9-37

Lab Answer Keys

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

About This Course xiii

About This Course This section provides you with a brief description of the course, audience, student prerequisites, course objectives, and a course outline.

Course Description This three-day instructor-led course introduces Windows Server®°2008 clustering, and provides students with the knowledge and skills to implement, maintain, and troubleshoot clusters.

Audience This course is intended for IT professional technical specialists who are responsible for using clustering technologies to implement and maintain high-availability solutions.

Student Prerequisites This course requires that you meet the following prerequisites:

• Experience managing Windows Server®°2003 or Windows Server 2008, in a networking environment

• Basic knowledge of clustering theory

• Experience managing applications and network topologies in an enterprise environment

• Basic IT troubleshooting skills

Course Objectives After completing this course, you will be able to:

• Provide an overview of cluster concepts.

• Describe key features and functionality of the Windows Server 2008 failover clusters.

• Prepare to install a failover cluster.

• Describe failover cluster storage requirements.

• Configure a failover cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

xiv About This Course

• Configure cluster resources and server roles.

• Maintain and troubleshoot failover clusters.

• Describe how to use and implement multi-site clusters.

• Explain how to install and maintain network load balanced (NLB) clusters.

Course Outline The following section provides a course outline:

Module 1: Provides an overview of cluster concepts and functionality.

Module 2: Describes key features and functionality of the Windows Server 2008 failover clusters.

Module 3: Explains the technical prerequisites and planning that is required to install a Windows failover cluster.

Module 4: Describes storage fundamentals, and how to plan and implement storage solutions for failover clusters.

Module 5: Explains how to install and manage a failover cluster.

Module 6: Explains how to configure cluster resources, and how to cluster common Window Server roles and applications.

Module 7: Explains how to maintain and troubleshoot failover clusters.

Module 8: Explains geographically dispersed clusters, and the challenges that they present. This module also describes how to implement a multi-subnet cluster using Windows Server 2008.

Module 9: Explains how to install and maintain network load balanced (NLB) clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

About This Course xv

Course Materials Your kit includes the following materials:

• Course Handbook. The handbook contains the materials that the class will cover.

• Course evaluation. At the end of the course, you will have the opportunity to complete an online evaluation to provide feedback about the course, training facility, and instructor.

To provide additional comments or feedback about the course, send e-mail to [email protected]. To inquire about the Microsoft Certification Program, send e-mail to [email protected].

Virtual Machine Environment This section provides information about setting up the classroom environment to support the course’s business scenario.

Virtual Machine Configuration In this course, you will use Microsoft® Virtual Server 2005 and the Microsoft Lab Launcher to perform the labs.

Important: At the end of each lab, you must close the virtual machine, and must not save any changes. To close a virtual machine without saving changes, perform the following steps for each running virtual machine: 1. Close the Virtual Machine Remote Control window. 2. In the Close box that appears, select Turn off machine and discard changes, and then click OK. 3. Close the Lab Launcher window.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

xvi About This Course

The following table shows the role of each virtual machine that this course uses:

Virtual machine Role

6423A-VAN-DC1 Domain controller in the Fabrikam.com domain

Windows Server 2003

6423A-VAN-SRV3A

6423A-VAN-SRV3B Member servers in Fabrikam.com

6423A-VAN-SRV5A


6423A-VAN-SRV6A


6423A-VAN-CORE6A

6423A-VAN-CORE6B Member servers in Fabrikam.com

Software Configuration The following software is installed on each virtual machine:

• Windows Server 2008 Enterprise

Classroom Setup Each classroom computer will have the same virtual machine configured in the same way.

Course Hardware Level To ensure a satisfactory student experience, Microsoft Learning requires a minimum equipment configuration for trainer and student computers in all Microsoft Certified Partner for Learning Solutions (CPLS) classrooms in which Official Microsoft Learning Product courseware is taught. This course requires a computer that exceeds hardware level 5.5, and should consist of a 2.4–gigahertz (minimum) Pentium 4 or equivalent CPU, at least 4 gigabytes (GB) of RAM, 16 megabytes (MB) of video RAM, and a 7200 RPM 80-GB hard disk.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Introduction to Clusters 1-1

Module 1 Introduction to Clusters

Contents: Lesson 1: Overview of Clusters 1-3

Lesson 2: Benefits of Using Clusters 1-11

Lesson 3: Overview of the Windows Server 2008 High-Availability Solutions 1-20

Lab: Identifying Windows Server 2008 High-Availability Solutions 1-29

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

1-2 Implementing and Managing Windows Server 2008 Clustering

Module Overview

Today’s enterprises place increased emphasis on high availability and performance of mission-critical applications. To find effective solutions for these environments, it is important to understand both industry standard terminology, and the available Windows Server®°2008 solutions.

This module describes clustering technologies, availability, scalability, and manageability features, and Windows Server 2008 cluster solutions.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Clusters

This lesson provides the industry definition of clusters, and describes how clusters can benefit an enterprise. This lesson also defines important cluster terminology, and the cluster types and categories.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Clusters?

Key Points A cluster is a group of computers and storage devices that work together as a single organized system. You can use clusters to distribute servicing load or provide high availability of services. You can create many different types of clusters to provide these services.

In a cluster, the component computers communicate with each other over a high-performance, reliable network: they may share one or more common storage devices. You use cluster configurations to address availability, scalability, and manageability.

For more information, see "What’s New in Clustering Technologies"

For more information, see "Failover Clustering with Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Why Use Clusters?

Key Points Answer the questions in a classroom discussion.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Clustering Terminology

Key Points You must have a basic understanding of common terminology to discuss cluster features and components.

For more information, see "Windows Server 2003 R2 Enterprise Edition: Cluster Server Resource Center"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Types of Clusters

Key Points There are four main types of clusters that are used in the market today. The four types are:

• Highly available (HA) clusters, which provide greater uptime for computing services. Multiple nodes allow backup servers to be available if a running server suffers a failure.

• Load-balanced clusters, which distribute incoming network load to multiple servers. Resources between servers are not usually shared. These types of clusters are primarily designed to spread the work of one server to multiple servers. Load-balanced clusters are also called server farms.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• High performance computing (HPC) clusters, which spread computational tasks among multiple nodes. HPC clusters are very tightly interconnected and require active communication between cluster servers. HPC clusters are able to take advantage of the parallel processing available with multiple nodes, but act like a single computational entity. HPC clusters are commonly used in the science, medical, and academic communities.

• Grid computing clusters are similar to HPC clusters; however, the nodes are not as tightly connected and do not function as a single computational entity. Tasks or jobs are farmed out to nodes that perform the work independently from the rest of the cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Clustering Categories

Key Points The four types of clusters fall into two main categories: shared device, and shared nothing clusters.

• In shared device clusters, software that runs on any node in the cluster may access any resource (for example, a disk) that is connected to any other node in the cluster. If two nodes must access the same data, the data must be either read twice from the disk, or copied from one node to another. You typically employ a locking and tracking mechanism to preserve data integrity.

• Shared nothing provide greater redundancy. Each node of the cluster owns a subset of the resources that make up the cluster. Only one system may own and access a particular resource at a time, although, on a failure, another node may take ownership of the resource. Additionally, requests from clients are automatically routed to the node that owns the resource.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Windows Server 2003 R2 Enterprise Edition: Cluster Server Resource Center"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Benefits of Using Clusters

Clusters provide availability, scalability, and manageability for mission-critical applications. This lesson describes how clusters benefit an enterprise.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Clustering Benefits

Key Points Clusters provide benefits in three key areas: availability, scalability, and manageability.

For more information, see "Windows Server 2003 Clustering"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Availability?

Key Points Availability is measured in the percentage of time that resources are running and accessible during the course of a year or other period. Generally, availability rates do not include planned downtime. Planned downtime includes time that the application is unavailable due to software, hardware, and other maintenance activities.

Mean time to failure (MTTF) is the amount of time that it takes for a solution to fail. Every component has an MTTF. For example, a hard disk might have an MTTF of 150,000 hours, and on average, you would replace it after 150,000 hours of use. To increase MTTF of a single hard disk, you can add multiple disks in a redundant array. You can apply this same methodology to other components, such as memory and power supplies.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Decreasing the mean time to recover (MTTR) of solution components also helps improve availability. You can help reduce the time it takes to recover from a failure by creating standard operating procedures and fully documenting a solution.

Although clusters can improve MTTF and MTTR by increasing the number of available nodes, it is also important to address the MTTF and MTTR of individual cluster components in order to improve overall cluster stability and performance.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Clusters Improve Availability

Key Points Clusters help improve:

• MTTF by spreading load across multiple servers. Redundant cluster nodes improve the MTTF by spreading load between nodes, which reduces the amount of time a resource is used.

• MTTR by having multiple nodes, which allows a cluster solution to respond as soon as a failure is detected.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Increasing the Mean Time to Failure


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Scalability?

Key Points Clusters can address several types of scalability: scaling up, scaling out, and consolidation. Consolidation may seem to be the opposite of scalability; however, due to improved stability and performance of clusters, you might be able to add more load to a cluster.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Clusters Improve Scalability

Key Points When scaling out, you add more nodes to a cluster to improve performance across all nodes. For example, suppose that you have a two-node cluster with both nodes running at 75% processor utilization, and the end-user performance is impacted. If you add another node to the cluster, you can reduce the processor utilization. This reduction occurs because the additional node takes on part of the workload. The result is improved performance for end users.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Cluster Manageability?

Key Points Although complexity increases in a clustered environment, disaster recovery and update management capabilities outweigh these complexities. For example:

• Disaster recovery improves in a cluster because application and hardware failures are recovered automatically by restarting the application on a healthy node. This enables the failed node to be repaired while the application remains available. In a stand-alone server configuration, you might need to reinstall or reconfigure an application on new hardware in the event of a failure.

• By using a cluster, you can perform application and operating system updates while the application continues to run on another cluster node.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Overview of the Windows Server 2008 High-Availability Solutions

Windows Server 2008 provides two primary high-availability solutions: failover clusters, and network load balancing. This lesson describes how each type works separately, and how they work together.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Windows Server 2008 Failover Cluster Features

Key Points You assign most clustered applications and their associated resources to one cluster node at a time. This type of resource control is commonly called a shared nothing model. In this model, only the owning node has access to the assigned cluster resources.

If a failover cluster detects a node failure or if the node is taken offline for maintenance while running a clustered application, the application is started on another cluster node. To minimize the impact of the failure, client requests are immediately redirected to the new cluster node.

For more information, see "Failover Clustering with Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Availability and Scalability"

For more information, see "Windows Server 2008 Failover Clustering Help: Overview of Failover Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Windows Server 2008 Failover Clusters Work

Key Points In a Windows Server 2008 failover cluster, only one node at a time controls resources. For example, in the case of a computer running Microsoft® SQL Server®, the SQL Server application runs on either node A or node B. Required SQL Server resources are controlled by the node that runs SQL Server. The node that controls the resources is called the active, node and the node that is not running any resources is called the passive node.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Windows Server 2008 Network Load Balancing (NLB) Clusters

Key Points You can use network load balancing clusters to provide scalability for Web services and other front-end servers such as virtual private network (VPN) servers and firewalls. Organizations can build groups of clustered computers to support load balancing of Transmission Control Protocol (TCP) and User Datagram Protocol (UDP) traffic requests.

Network load balancing clusters are groups of identical (usually cloned) computers that, because of their numbers, enhance the availability of:

• Web servers

• Microsoft Internet Security and Acceleration (ISA) servers (for proxy and firewall servers)

• Terminal Services

• Microsoft Office Outlook® Web Access

• Other applications that receive TCP and UDP traffic

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Because network load balancing cluster nodes are usually identical clones of each other, and can therefore operate independently, all nodes in a network load balancing cluster are active.

You can scale out network load balancing clusters by adding as many as 32 servers. Each node runs a copy of the IP-based application or service that is being load-balanced. Each node has access to all the data that is necessary for the application or service to run.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Network Load Balancing Clusters Work

Key Points Network load balancing distributes IP traffic to multiple instances of a TCP/IP service, such as a Web server, each running on a host within the cluster. Network load balancing transparently distributes the client requests among the hosts, and enables the clients to access the cluster using one or more virtual IP addresses. From the client’s point of view, the cluster appears to be a single server that answers these client requests. As enterprise traffic increases, you can simply plug another server into the cluster.

For example, clustered hosts work together to service network traffic from the Internet. Each server runs a copy of an IP-based service, such as Internet Information Services 7.0 (IIS), and network load balancing distributes the workload among them. This speeds up normal processing so that Internet clients receive faster turnaround on their requests.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Summary of Network Load Balancing Clusters

Key Points Network load balancing is usually used to provide high availability and scalability for Web servers or File Transfer Protocol (FTP) servers. No special hardware is required to use network load balancing, and it is available in all editions of Windows Server 2008. Network load balancing is usually deployed in a single network, but can also be deployed over multiple networks to provide cross-network availability.

For more information, see "Network Load Balancing Improvements"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Failover Clusters with Network Load Balancing

Key Points You can often use failover clusters and network load balancing together to create a comprehensive solution. For example, a Web application can have a network load-balanced tier of Web servers that connect to a failover clustered database tier. This type of highly available tiered configuration provides a redundant solution that can sustain multiple failures while continuing to serve client requests.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Identifying Windows Server 2008 High-Availability Solutions

Lab Objectives • Identify solutions for Web servers

• Identify solutions for database servers

• Identify complex solutions

Scenario In this lab, you will answer questions based on given scenarios.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Identifying Solutions for Web Servers Fourth Coffee has a company Web site that contains mostly static content. The number of users that connect to the site continues to increase. The site runs on a single Windows Server 2008 computer that cannot adequately handle the peak load.

The main task for this exercise is to identify solutions based on the given scenario.

Task 1: Identify solutions based on the given scenario 1. What technology can you use to provide more processing power to the Web

site?

2. How would this technology solve the problem of a single point of failure?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Identifying Solutions for Database Servers Contoso has a single computer running Microsoft SQL Server, whose job it is to process orders. Because of monthly maintenance routines and hardware failures, the server is unavailable for several hours every month. Contoso needs a solution that minimizes downtime during normal updates, and minimizes outages because of hardware failures.


Task 1: Identify solutions based on the given scenario 1. What technology can you use to minimize downtime during operating system

updates and hardware failures?

2. How would this solution solve the problem of a single point of failure?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 3: Identifying Complex Solutions Contoso has a database-driven Web site with a single server. Performance and server hardware problems affect the amount of orders that can be placed with the site. Contoso needs a solution that improves the order processing site’s performance and availability.


Task 1: Identify solutions based on the given scenario 1. What clustering solution(s) can you use to solve the Web site problems?

2. At a minimum, how many servers would you need to complete this configuration?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Introduction to Windows Server 2008 Failover Clusters 2-1

Module 2 Introduction to Windows Server 2008 Failover Clusters

Contents: Lesson 1: Overview of Windows Server 2008 Failover Clusters 2-3

Lesson 2: Key Windows Server 2008 Failover Cluster Features 2-12

Lesson 3: Overview of the Windows Server 2008 Quorum Modes 2-23

Lab: Identifying Windows Server 2008 Clustering Solutions 2-36

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

Failover clusters in Windows Server® 2008 provide a high-availability solution for many server roles and applications. By implementing failover clusters, you can maintain application or service availability if one or more computers in the failover cluster fail.

This module provides an overview of Windows Server 2008 failover clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Windows Server 2008 Failover Clusters

In order to plan and implement a failover cluster, you must first understand the terminology, and how failover clusters work. This lesson provides this information.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Components

Key Points In a failover cluster, each node in the cluster:

• Has full connectivity and communication with the other nodes in the cluster.

• Is aware when another node joins or leaves the cluster.

• Is connected to a network through which client computers can access the cluster.

• Is connected through a shared bus or Internet SCSI (iSCSI) connection to shared storage.

• Is aware of the services or applications that are running locally, and the resources that are running on all other cluster nodes.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Most clustered applications and their associated resources are assigned to one cluster node at a time. The node that has access to those cluster resources is the active node. If the nodes detect the failure of the active node for a clustered application, or if the active node is taken offline for maintenance, then the clustered application is started on another cluster node. To minimize the impact of the failure, client requests are immediately and transparently redirected to the new cluster node.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Components Terminology

Key Points Shared storage usually refers to logical devices — normally hard drives or logical unit numbers (LUN) — that all the cluster nodes attach to via a shared bus. This bus is separate from the bus that contains the system and boot disks. The shared disks store applications, file shares, and other resources that the cluster will manage.

For more information, see "Step-by-Step Guide for Configuring a Two-Node File Server Failover Cluster in Windows Server 2008"

For more information, see "What’s New in Failover Clusters in Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Clustered Services and Resources?

Key Points Clustered services that contain an IP address resource and a network name resource (as well as other resources,) are published to clients on the network under a unique server name. Because these groups appear as individual logical servers to clients, they are called a cluster instance.

Users access applications or services on an instance in the same way they would if the application or service were on a non-clustered server. Usually, the application or user does not know that they are connecting to a cluster, nor to which node they are connected.

Resources are physical or logical entities, such as a file share, disk, or IP address, which the failover cluster manages. Resources may provide a service to clients, or be an integral part of the cluster. Resources are the most basic and smallest configurable unit. At any given time, a resource can run only on a single node in a cluster, and is online on a node when it is providing its service to that specific node.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Clusters and Networks

Key Points When configuring networks in failover clusters, you must dedicate a network to connect to the shared storage. If you use iSCSI for the shared storage connection, the network will use IP-based communications; however, you cannot use this network for node or client communication.

You can use the private and public networks for both client and node communications. As a best practice, you should dedicate an isolated network for the private node communication. The public network is configured to allow client connections to the failover cluster. The public network can provide backup for the private network.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Step-by-Step Guide for Configuring a Two-Node File Server Failover Cluster in Windows Server 2008"


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Failover?

Key Points A failover attempt consists of the following steps:

1. The cluster service takes all the resources in the instance offline in an order that is determined by the instance’s dependency hierarchy: dependent resources first, followed by the resources on which they depend. For example, if an application depends on a physical disk resource, the cluster service takes the application offline first, which allows the application to write changes to the disk before the disk is taken offline.

2. When all the resources are offline, the cluster service attempts to transfer the instance to the node that is listed next on the instance’s list of preferred owners.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


3. If the cluster service successfully moves the instance to another node, it attempts to bring all the resources online. This time it starts at the bottom of the dependency hierarchy. Failover is complete when all of the resources are online on the new node.

4. When the node becomes active again, the cluster service can fail back the instances that were originally hosted by the node. When the cluster service fails back an instance, it uses the same procedures that it performs during failover: that is, the cluster service takes all the resources in the instance offline, moves the instance, and then brings all the resources in the instance back online.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Key Windows Server 2008 Failover Cluster Features

Windows Server 2008 includes several core features that make it easier to set up and maintain a failover cluster. This lesson describes these features and how they will affect your failover administration tasks.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Installation and Validation Features

Key Points One of the significant new features in Windows Server 2008 failover clusters is that all the hardware components must be marked as “Certified for Windows Server 2008”. In addition, the complete configuration must pass all tests in the Validate a Configuration Wizard. This is different from previous server clusters versions, which required that all hardware be on a hardware compatibility list (HCL), specifically a list for server clusters.

The Validate a Configuration Wizard tests include specific simulations of cluster actions, which fall into the following categories:

• System configuration tests: Analyzes whether the selected servers meet specific requirements, for example the requirement that the servers must run the same operating system version and software updates.

• Network tests: Analyzes whether the cluster nodes network configuration meets specific requirements, for example requirements for network redundancy.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Storage tests: Analyzes whether the storage meets specific requirements, for example whether the storage correctly supports the necessary SCSI commands and handles simulated cluster actions correctly.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Management Features

Key Points The management tools install by default when you install the Failover Clustering feature. You can also install these tools on computers running Windows Server 2008 that do not have the Failover Clustering feature installed.

To install the management tools, choose the option to add a feature in Server Manager. The failover cluster tools are available under the Remote Server Administration Tools.

The failover cluster tools are also included in Remote Server Administration Tools that you can install on computers running Windows Vista® Service Pack 1.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Using the Failover Cluster Management Interface

Question:

How would you compare the Failover Cluster Management snap-in in Windows Server 2008 with server cluster management tools in previous Windows versions, or with other Windows Server 2008 management tools?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Networking Features

Key Points Networking features include the following features:

• In order to verify all nodes in a cluster, the nodes transmit and receive heartbeats to confirm that each node is still available. The nodes use User Datagram Protocol (UDP) unicast, versus UPD broadcast (used in legacy clusters). The messages are sent on port 3343.

• You can now place clustered servers on different IP subnets, which reduces the requirements for multi-site clusters.

• The Microsoft® Failover Cluster Virtual Adapter is a hidden device that is added to each node when you install the Failover Clustering feature. The adapter is assigned a media access control (MAC) address based of the MAC address that is associated with the first enumerated physical network adapter in the node.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Failover clusters fully support IP version 6 (IPv6) for both node-to-node and node-to-client communication.

• You can use Dynamic Host Configuration Protocol (DHCP) to assign IP addresses, or you can assign static IP addresses to all nodes in the cluster. If some nodes have static IP addresses, and you configure others to use DHCP, the Validate a Configuration Wizard will raise an error. The cluster IP address resources are obtained based on the configuration of the network interface supporting that cluster network.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Storage Features

Key Points Storage features include the following features:

• Parallel SCSI, which previous versions of server clusters supported, is no longer supported for shared disk devices in a failover cluster.

• In Windows Server 2008, the cluster disk driver is a stand-alone component that communicates with the Partition Manager driver (PartMgr.sys), for managing clustered disks. During the installation process, the driver is registered as a system Plug-n-Play (PNP) device, and is started during the normal Windows startup process.

The cluster disk driver provides improvements to disk fencing and arbitration behavior. Disk fencing activities have been offloaded to the PartMgr.sys driver. This change improves integration with core operating system disk management, and reduces the chances for disk corruption. Disk arbitration now uses persistent reservations and an improved algorithm. This works to eliminate the possibility of the disk being in an unprotected state.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Failover clusters support both Master Boot Record (MBR) and GUID Partition Table (GPT) disks. GPT disks allow you to create partitions greater than 2 terabytes. Dynamic disks are not supported.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Security Features

Key Points Security features include the following features:

• In Windows Server 2008 failover clusters, authentication uses Kerberos authentication, but it can fall back to NTLM authentication (version 1 or version 2), if needed. Because computer objects are created in Active Directory® domain services (AD DS), Kerberos authentication is supported by default, for each clustered Network Name resource.

• When you configure a failover cluster, you create both a Cluster Name Object (CNO) and a Virtual Computer Object (VCO). The CNO is associated with the Cluster Name core resource. A VCO is a computer object that is created in AD DS for all other cluster Network Name resources that are created as part of configuring a Client Access Point (CAP). After AD DS creates the CNO, it is used as the security context for the cluster. It is used when accessing all external resources, and in the creation of VCOs. As a result, Kerberos will be enabled on the Network Name resource by default, and will be used as the primary authentication mechanism.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• The Windows Server 2008 Failover Clustering auditing feature provides the capability to audit (monitor) cluster access. You can enable the auditing feature to audit accesses (success and\or failure) of the Microsoft Failover Cluster object. This means that any client access that uses either the Failover Cluster Management snap-in, the cluster.exe command line, or any custom application that calls cluster application programming interfaces (APIs), will be audited. The auditing feature only audits cluster access; actions executed once granted access to the cluster are not audited.

• By default, all communication between nodes is digitally signed, which verifies the source of all communication. You can use the cluster.exe command-line tool to set the SecurityLevel attribute for the cluster value of 2. This will encrypt inter-cluster communications and increase security.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Overview of the Windows Server 2008 Quorum Modes

The most important reason for implementing failover clusters is to provide high availability for services and applications. Windows Server 2008 failover clusters use a quorum to ensure that enough members are available for the cluster to continue operating. In order to implement and manage failover clusters, you must understand the options that are available for configuring quorums. This lesson describes how quorum works in Windows Server 2008 failover clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Quorum?

Key Points The failover cluster quorum configuration determines the number of failed nodes or failed storage and network components that the cluster can sustain while continuing to function. Quorum prevents two sets of nodes from operating simultaneously as the failover cluster. Simultaneous operation could happen when network problems prevent one set of nodes from communicating with another set of nodes. Without a quorum mechanism, each set of nodes could continue to operate as a failover cluster, resulting in a partition within the cluster.

To prevent problems caused by a split in the cluster, failover clusters use a voting algorithm to determine whether the cluster has enough votes to maintain quorum. Because a given cluster has a specific set of nodes and a specific quorum configuration, the cluster will know how many votes are required. If the number of votes drops below the majority, the cluster stops running. Nodes will still listen for the presence of other nodes, in case another node appears again on the network, but the nodes will not function as a cluster until a consensus is reached.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Online Help: Understanding Quorum Configurations in a Failover Cluster"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Quorum Works

Key Points The Quorum Manager component determines if the current cluster membership has quorum. If the Quorum Manager determines that quorum has not been achieved or has been lost, it will terminate the cluster across all nodes.

The Resource Control Manager component controls the configuration and state of resources, and resource dependency trees. It is responsible for monitoring the online status of active resources.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Types of Quorum Modes

Key Points With the exception of the No Majority: Disk Only mode, all quorum modes in Windows Server 2008 failover clusters are based on a simple majority vote model. As long as a majority of the votes are available, the cluster will continue to function. For example, if there are five votes in the cluster, the cluster will continue to function as long as there are at least three available votes. The source of the votes is not relevant—the vote could be a node, witness disk, or witness file share. When only a minority of votes is available, the cluster will stop functioning.

In the No Majority: Disk Only mode, the quorum-shared disk can veto all other possible votes. In this mode, the cluster will continue to function as long as the quorum-shared disk and at least one node are available. If the quorum-shared disk is not available, the cluster will stop functioning, even if all nodes are still available. In this mode, the quorum-shared disk is a single point of failure, so this mode is not recommended.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Online Help: Illustrations of Quorum Configurations"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the Node Majority Mode?

Key Points In a Node Majority mode, only the nodes in the cluster have a vote. If there are a total of three nodes in the cluster, the following scenarios are possible.

Node1 Node2 Node3 Cluster status

Online Online Online Online

Online Online Offline Online

Online Offline Offline Offline

Offline Offline Offline Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the Node and Disk Majority Mode?

Key Points In a Node and Disk Majority mode, both the nodes in the cluster and the witness disk have a vote. If there are a total of four nodes in the cluster, the following scenarios are possible.

Node1 Node2 Node3 Node4 Witness disk Cluster status

Online Online Online Online Online Online

Online Online Offline Online Offline Online

Online Online Offline Offline Online Online

Online Online Offline Offline Online Online

Online Online Offline Offline Offline Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Node1 Node2 Node3 Node4 Witness disk Cluster status

Online Offline Offline Offline Online Offline

Offline Offline Offline Offline Online Offline

Offline Offline Offline Offline Offline Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the Node and File Share Majority Mode?

Key Points In a Node and File Share Majority mode, the nodes in the cluster and a file share witness have a vote. If there are a total of three nodes in the cluster, the following scenarios are possible.

Node1 Node2 Node3 File share witness Cluster status

Online Online Online Online Online

Online Online Online Offline Online

Online Online Offline Offline Offline

Online Offline Offline Online Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Node1 Node2 Node3 File share witness Cluster status

Offline Offline Offline Online Offline

Offline Offline Offline Offline Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the No Majority: Disk Only Mode?

Key Points In a No Majority: Disk Only mode, only the shared disk has a vote. If there are a total of three nodes in the cluster, the following scenarios are possible.

Node1 Node2 Node3 Shared disk Cluster status

Online Online Online Online Online

Online Online Online Offline Offline

Online Online Offline Offline Offline

Online Offline Offline Online Online

Offline Offline Offline Online Offline

Offline Offline Offline Offline Offline

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Choosing a Quorum Mode

Key Points When you configure a failover cluster in Windows Server 2008, the Installation Wizard automatically chooses one of two possible default configurations. Modify this setting only if you determine that a change is appropriate for your cluster, and ensure that you understand the implications of making the change.

In addition to planning your quorum mode, you should also consider the capacity of the nodes in your cluster, and their ability to support the services and applications that may fail over to that node. For example, a cluster that has four nodes and a witness disk will still have quorum after two nodes fail. However, if you have several applications or services deployed on the cluster, each remaining cluster node may not have the capacity to provide services.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Identifying Windows Server 2008 Clustering Solutions

Lab Objectives • Identify clustered scenarios

Scenario In this lab, you will answer questions based on the following scenarios.

Scenario 1

Contoso is redesigning their file server implementation. The organization has identified high availability as an essential requirement in the new server implementation. All of the file servers will be located in the organization’s data center, and will be connected to a storage area network (SAN). To meet performance requirements, the organization will require three servers to be available. The organization has been using a SAN for several years.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Scenario 2

Tailspin Toys is redesigning their print server implementation. The organization has identified high availability as an essential requirement in the new server implementation. The print servers are deployed in one of three data centers. All data centers are connected by a wide area network (WAN), which is highly available and has excess bandwidth. To meet high availability requirements, printing must not be disrupted if the local print server that the user is connected to fails. The organization has implemented a third-party data replication product between the data centers.

Scenario 3

Woodgrove Bank is planning to deploy a new failover cluster-aware application. The application provides essential business services, and must be highly available. All of the application servers will be deployed in one data center, and all servers will be connected to a SAN using Internet SCSI (iSCSI). The application is highly disk input/output (I/O)-intensive, so the implementation should minimize disk I/O as much as possible.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Identifying Clustered Scenarios The main task for this exercise is to identify cluster solutions based on the given scenarios.

Task 1: Identify cluster solutions based on each of the previous scenarios 1. Which quorum mode would you recommend for each scenario?

2. How would you deploy and configure the nodes in each of the scenarios?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Preparing to Install a Failover Cluster 3-1

Module 3 Preparing to Install a Failover Cluster

Contents: Lesson 1: Overview of Requirements for Installing a Failover Cluster 3-3

Lesson 2: Planning the Failover Cluster Implementation 3-13

Lesson 3: Installing the Failover Clustering Feature and Validating the Cluster Configuration 3-26

Lesson 4: Installing the Failover Cluster on Windows Server 2008 Server Core 3-41

Lab: Preparing for a Cluster Installation 3-47

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

This module outlines the requirements for failover cluster hardware and software. Additionally, this module outlines steps to install the Failover Clustering feature, use the Validate a Configuration Wizard to ensure proper cluster configuration, and install the failover cluster on Windows Server® 2008 Server Core.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Requirements for Installing a Failover Cluster

Windows Server 2008 failover clusters have specific recommended hardware and software configurations that enable Microsoft® to support the cluster. Failover clusters are intended to provide a higher level of service than stand-alone servers. Therefore, cluster hardware requirements are often more strict than requirements for stand-alone servers.

This lesson describes the hardware, network, storage, infrastructure, and software requirements for Windows Server 2008 failover clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Server Hardware Requirements

Key Points Failover clusters need the following to meet availability and support requirements:

• All hardware that you select for a failover cluster must meet the “Certified for Windows Server 2008” logo requirements. Hardware that obtains this logo is independently tested to meet the highest technical bar for reliability, availability, stability, security, and platform compatibility.

• You must install the same hardware on each failover cluster node. For example, if you choose a specific model of network adapter, you should install this adapter on each of the cluster nodes.

• After you configure the servers with the hardware, all Validate a Configuration Wizard tests must pass before the cluster is considered a configuration that is supported by Microsoft.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Help: Understanding Requirement for Failover Clusters"

For more information, see "Windows Server 2008 Application Compatibility and Certification Forum"

For more information, see "Innovate on Windows Server"

For more information, see "Windows Server Catalog of Tested Products"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Server Network Requirements

Key Points Failover cluster network components must have both the “Certified for Windows Server 2008” logo, and pass the Validate a Configuration Wizard tests. Additionally:

• The network adapters in each of the nodes must be identical and have the same IP protocol version, speed, duplex, and flow control capabilities available.

• The networks to which you connect the nodes should be redundant so that even a single failure allows the nodes to continue to communicate with each other. You can use network adapter teaming to provide single network redundancy. We recommend multiple networks to provide multiple paths between nodes for inter-node communication; otherwise, a warning will be generated during the validation process.

• All network adapters in a cluster network must have the same IP address format, which means they all use static IP addresses, or all use Dynamic Host Configuration Protocol (DHCP).

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Help: Understanding Requirements for Failover Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Storage Requirements

Key Points Storage is one of the most critical components in a failover cluster and requires that:

• The mass-storage controllers match in each server node, in model, and in firmware versions.

• The drivers are based on the Microsoft Storport driver model, and that all Multipath software must be based on Microsoft Multipath I/O (MPIO). MPIO can use generic device modules that the operating system provides, or device-specific modules that the hardware vendor develops to work with the hardware.

• Internet SCSI-based (iSCSI-based) storage has a dedicated network that is separate from the client access and cluster-only communications networks.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Windows Server provides two disk storage types: basic and dynamic. A single node owns dynamic disks, which are not able to fail over. Dynamic disks are not supported in failover clusters natively. Failover clusters only support basic disks, and will support both Master Boot Record (MBR) and GUID Partition Table (GPT) disk partitioning systems. Because Microsoft supports both the NTFS file system and the FAT32 file system, you can format the clustered disks with either system. We recommend that you format the clustered disks with the NTFS file system, which is also a requirement for all File Share Witness volumes.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Infrastructure Requirements

Key Points Failover clusters depend on infrastructure services, and require that each server node must be in the same Active Directory® domain and, if you use DNS, the nodes should use the same Domain Name System (DNS) servers for name resolution.

We recommend that you install the same Windows Server 2008 features and roles on each node. In addition, you should not install the Active Directory Domain Services (AD DS) role on any of the cluster nodes. If you install the AD DS role on one of the nodes, you must install it on all nodes.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Software Requirements

Key Points Failover clusters require that each cluster node run the same edition of either Windows Server 2008 Enterprise, or Windows Server 2008 Datacenter. Depending on the role that will be clustered, Server Core edition may also meet the software requirements. However, you cannot install Server Core and full editions in the same cluster.

Each node must run the same processor architecture. This means that each node will have the same processor family, which might be Intel Xeon processor with Extended Memory 64Technology®, AMD Opteron AMD64, or Intel Itanium-based.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Guidelines for Microsoft Support of Cluster Solutions

Key Points The new support policy is different from the support policy for previous Windows Server versions. The new policy uses the Validate a Configuration Wizard to provide detailed configuration tests of the failover cluster. This ensures that the cluster matches best practices, and is configured in a way that Microsoft can support.

For more information, see "The Microsoft Support Policy for Windows Server 2008 Failover Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Planning the Failover Cluster Implementation

After you determine the requirements for a failover cluster, you must verify that the required components are in place. This lesson outlines additional steps that you must take before creating a cluster. These additional steps will help ensure a stable and successful implementation.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Preparing the Server Hardware Before Validating the Cluster

Key Points Before you run the Validate a Configuration Wizard, review all cluster requirements and verify that:

• All hardware and software is in place and properly configured. Windows Failover Cluster Help provides step-by-step instructions to assist you with preparing hardware.

• You have a plan for each clustered application, which ensures that the cluster is configured to meet business needs.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED



For more information, see "Windows Server 2008 home page"

For more information, see "Failover Cluster Help: Prepare Hardware Before Validating a Failover Cluster"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Understanding the Cluster Security Model

Key Points Windows Server 2008 creates a Clustered Name Object (CNO) during the cluster creation process. The CNO is used to access resources such as AD DS, DNS, and Windows Internet Name Service (WINS), and for creating cluster Network Name resources. By default, in Windows Server 2008, clustered applications take advantage of Kerberos authentication for resources. Windows Server 2008 failover clusters provide support for legacy authentication, and for other applications that require NT LAN Manager (NTLM) authentication.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Identifying Single Points of Failure


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Addressing Single Points of Failure

Key Points You must look at all cluster configuration components to identify single points of failure. You can remediate many single points of failure with simple solutions, such as adding storage controllers, or network adapters, and multipathing software. These solutions will reduce the probability that a device failure will cause a failure in the cluster. Typically, server class hardware has options for multiple power supplies for power redundancy, and for creating redundant array of independent disks (RAID) sets for disk data redundancy.

Some single points of failures might be difficult or expensive to solve; in these situations, you must weigh the benefits of eliminating the point of failure against the cost of purchasing and maintaining the solution. For example, if a single datacenter fails, creating a multi-site cluster can solve this problem; however, this solution is more complicated and expensive than a single location cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Operating System Upgrades

Key Points Windows Server 2008 cluster service has changed from previous versions. It does not allow legacy clusters to exist, or to be upgraded to Windows Server 2008. However, the Migrate Services and Applications Wizard can help migrate specific resources from legacy clusters to new Windows Server 2008 clusters. This wizard will migrate the following resources:

• Physical Disk resource

• Network Name resource

• IP Address resource

• File Share resource (including DFS Root)

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The wizard will also migrate configuration information for the following resources.

• DHCP resource

• WINS resource

• Generic Application resource

• Generic Service resource

• Generic Script resource

However, you will need to manually complete additional configuration steps for the migration, such as installing supporting server roles and copying data.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Planning for Migration

Key Points To migrate resources:

1. In the Failover Cluster Management snap-in, select or specify the cluster to which you want to migrate settings.

2. Under Configure, click Migrate Services and Applications.

3. Read the first page of the Migrate a Cluster Wizard, and then click Next.

4. Specify the name or IP Address of the cluster or cluster node from which you want to migrate resource groups, and then click Next.

5. Click View Report. Read the pre-migration report, which explains whether each resource is eligible for migration, and describes additional steps to perform after running the wizard.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


6. Follow the instructions in the wizard to complete the following:

• Choose the resource group or groups whose settings you want to migrate.

• Specify whether the resource groups to be migrated will use new storage, or the same storage used in the old cluster. If the resource groups will use new storage, you can specify the disk that each resource group should use after migration.

7. When the Summary page appears, click View Report. This report contains important information about any additional steps you must perform before you bring the migrated resource groups online. For example, you might need to copy applications from the old cluster to the new cluster.


For more information, see "Step-by-Step Guide for Migrating Settings from a Cluster Running Windows Server 2003 to a Cluster Running Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Server Capacity Planning

Key Points There are two main cluster load types: Active/Active, and Active/Passive. In an Active/Active configuration, each node hosts clustered services. In an Active/Passive configuration, at least one cluster node does not host any services.

Planning server capacity in an Active/Active configuration requires that when a node fails, the service will be able to continue functioning on an operational node that is running other services.

Planning server capacity in an Active/Passive configuration is dependent on the number of failed nodes the cluster is able to sustain.

When planning server capacity, you will determine the fewest number of nodes required to maintain the cluster operations during a failure, and to support the applications and hardware in the cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Planning for a Failed Node

Key Points If a node failure occurs in a two-node cluster with each node running at thirty percent of capacity, a single node must be able to handle the additional load from both services. To plan for this additional load, you must be sure that when a failover does occur, both services are maintained. If you do not plan for the additional load, and the load was sixty percent on each node in a two-node cluster, a failover would impact both clustered applications. In contrast to a stand-alone situation, a failure on a two-node cluster affects both the application originally running on the failed node, and the application running on the active node.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Installing the Failover Clustering Feature and Validating the Cluster Configuration

This lesson describes how to install the Failover Clustering feature, run the Validate a Configuration Wizard, and correct errors that the Validate a Configuration Wizard finds.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Process of Installing the Failover Clustering Feature

Key Points After you configure the hardware and have identified and addressed any single points of failure, you can begin to configure the cluster. To do this, install the Failover Clustering feature on each server node, and then run the Validate a Configuration Wizard to be sure that hardware and software meets the cluster requirements. Finally, create the cluster and configure highly available applications and services.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the Validate a Configuration Wizard?

Key Points The Validate a Configuration Wizard runs tests that confirm that the hardware and hardware settings are compatible with failover clustering. Using the wizard, you can run the entire set of configuration tests or a subset of the tests. We recommend that you run the tests on servers and storage devices before you configure the failover cluster, and again after any major changes are made to the cluster. You can access the test results in the %windir%\cluster\Reports directory.

The report from the Validate a Configuration Wizard includes one of the following outcomes for each test:

• The tested item meets requirements for a failover cluster.

• The tested item might meet the requirements, but you receive a warning notice suggesting that you review the hardware and software settings to be sure that best practices are met.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• The test item does not meet requirements.

• The test could not be run. For example, if you are testing only a single node, certain tests do not run.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Cluster Validation Test: Inventory

Key Points Inventory tests provide lists of information about the hardware, software, and settings for each server that you test. For example, after you perform scheduled maintenance, you can use the inventory test to ensure that the software updates on each server are identical. The inventory tests can be used separate from the other tests included in the Validate a Configuration Wizard.

For more information, see "Failover Cluster Help: Understanding Cluster Validation Tests: Inventory"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Cluster Validation Test: Network

Key Points The Validate a Configuration Wizard performs the following network tests:

Test Description

Validate Cluster Network Configuration

• Lists the cluster networks, which is the network topology as seen from the perspective of the cluster.

• Validates for a particular cluster network, that all network adapters are provided with IP addresses in the same way. (All use static IP addresses or all use DHCP.)

• Validates for a particular cluster network, that all network adapters use the same IP version. (All use IPv4, all use IPv6, or all use both IPv4 and IPv6.)

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Test Description

Validate IP Configuration

• Lists the IP configuration details.

• Validates that IP addresses are unique in the cluster, and that there is no duplication.

• Checks the number of network adapters on each tested server. If only one adapter is found, the report provides a warning about avoiding single points of failure in the network infrastructure that connects clustered servers. Single points of failure can be avoided in multiple ways: you can connect your cluster nodes by multiple, distinct networks, or alternatively, you can connect your cluster nodes with one network that is constructed with teamed network adapters, redundant switches, redundant routers, or similar hardware that removes single points of failure.

• Validates that no tested servers have multiple adapters on the same IP subnet.

• Validates that all tested servers use the same IP version. (All use IPv4, all use IPv6, or all use both IPv4 and IPv6.)

Validate Network Communication

• Validates that tested servers can communicate with acceptable latency on all networks. The test also checks to see if there are redundant communication paths between all servers. Communication between the nodes of a cluster enables the cluster to detect node failures and status changes, and to manage the cluster as a single entity.

Validate Windows Firewall Configuration

• Validates that Windows Firewall is configured correctly for failover clustering on the tested servers.

For more information, see "Failover Cluster Help: Understanding Cluster Validation Tests: Network"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Cluster Validation Test: Storage

Key Points The storage tests list and test the capabilities of all disks available to the cluster. These tests are comprehensive; however, some specific tests may not run after the cluster is running nor in a multi-site cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Validate a Configuration Wizard performs the following storage validation tests:

Test Description

List All Disks

Lists all disks that are visible to one or more tested servers. The test lists:

• Disks that can support clustering and can be accessed by all the servers.

• Disks on an individual server.

List Potential Cluster Disks

Lists disks that can support clustering, and are visible to all tested servers. To support clustering, the disk must be connected through Serial Attached SCSI (SAS), iSCSI, or Fibre Channel. In addition, the test validates that multipath I/O is working correctly, which means that each of the disks is seen as one disk, not two.

Validate Disk Access Latency

Validates that the latency for disk read/write operations is within an acceptable limit for a failover cluster. If disk read/write operations take too long, one possible result is that cluster time-outs might be triggered. Another possible result is that the application attempting to access the disk might appear to have failed, and the cluster might initiate a needless failover.

Validate Disk Arbitration

Validates that:

• Each of the clustered servers can use the arbitration process to become the owner of each of the cluster disks.

• When a particular server owns a disk, if one or more other servers arbitrate for that disk, the original owner retains ownership.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Test Description

Validate Disk Failover

Validates that disk failover works correctly in the cluster. Specifically, the test validates that when a disk owned by a clustered server is failed over, the server that takes ownership of the disk can read it. The test also validates that information written to the disk before the failover, is still the same after the failover.

If disk failover occurs but the server that takes ownership of a disk cannot read it, the cluster cannot maintain availability of the disk. If information written to the disk is changed during the process of failover, it might cause issues for users or software that require this information. In either case, if the affected disk is a witness disk, (a disk that stores cluster configuration data and participates in quorum,) such issues could cause the cluster to lose quorum and shut down.

Validate File System Validates that the file system on disks in shared storage is supported by failover clusters.

Validate Microsoft MPIO-Based Disks

Validates that multi-path disks (Microsoft Multipath I/O-based disks,) have been configured correctly for failover cluster.

Validate Multiple Arbitration

Validates that when multiple clustered servers arbitrate for a cluster disk, only one server obtains ownership.

If multiple clustered servers obtain ownership of a cluster disk through disk arbitration, the disk might become corrupted. Failover clusters are designed to operate in circumstances where only one clustered server at a time owns a disk. If multiple servers own a disk at the same time, they might perform write operations in an uncoordinated way, possibly corrupting the disk.

Validate SCSI Device Vital Product Data (VPD)

Validates that the storage supports necessary SCSI inquiry data as well as Vital Product Data (VPD) descriptors, and that they are unique.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Test Description

Validate SCSI-3 Persistent Reservation

Validates that the cluster storage uses the more recent (SCSI-3 standard) Persistent Reserve commands, which are different from the older SCSI-2 standard reserve/release commands. Because the Persistent Reserve commands avoid SCSI bus resets, they are much less disruptive than the older reserve/release commands. Therefore, a failover cluster can be more responsive in a variety of situations, unlike a cluster running an earlier version of the operating system. In addition, disks are never left in an unprotected state, which lowers the risk of volume corruption.

Validate Simultaneous Failover

Validates that simultaneous disk failovers work correctly in the cluster. Specifically, the test validates that even when multiple disk failovers occur simultaneously, any clustered server that takes ownership of a disk can read it. The test also validates that information written to each disk before a failover is the same after the failover.

If disk failover occurs but the server that takes ownership of a disk cannot read it, the cluster cannot maintain availability of the disk. If information that is written to the disk is changed during the process of failover, it might cause issues for users or software that requires this information. In either case, if the affected disk is a witness disk, (a disk that stores cluster configuration data and participates in quorum,) such issues might cause the cluster to lose quorum and shut down.

For more information, see "Failover Cluster Help: Understanding Cluster Validation Tests: Storage"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Cluster Validation Test: System Configuration

Key Points System configuration tests analyze selected servers to determine whether they are properly configured to work together in a failover cluster.

The Validate a Configuration Wizard performs the following system configuration tests:

Test Description

Validate Active Directory Configuration

Validates that each tested server is in the same domain and organizational unit. It also validates that all tested servers are domain controllers, or are member servers. To change the domain role of a server, use the Server Manager.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Test Description

Validate All Drivers Signed

Validates that all tested servers contain only signed drivers. If an unsigned driver is detected, the test is not considered a failure, but a warning is issued.

Signed drivers indicate whether the drivers on your system are original, unaltered files that either came with the operating system, or were supplied by a vendor.

Validate Operating System Versions

Validates that all tested servers have the same operating system version.

Validate Required Services

Validates that the services required for failover clustering are running on each tested server, and are configured to start automatically whenever the server is restarted.

Validate Same Processor Architecture

Validates that all tested servers have the same architecture. A failover cluster is supported only if the systems in it are all 32-bit systems, or all 64-bit systems.

Validate Service Pack Levels

Validates that all tested servers have the same service packs. A failover cluster can run even if some servers have service packs that are different than others. However, servers with different service packs might behave differently from each other, with unexpected results. We recommend that all servers in the failover cluster have the same service packs.

Validate Software Update Levels

Validates that all tested servers have the same software updates. A failover cluster can run even if some servers have different updates than others. However, servers with different software updates might behave differently from each other, with unexpected results. We recommend that all servers in the failover cluster have the same software update levels.

For more information, see "Failover Cluster Help: Understanding Cluster Validation Tests: System Configuration"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Correcting Cluster Validation Errors

Key Points The Validate a Configuration report contains details about the tests that were run, and their test results. The report also provides information about reasons for any failures. You can also review the Failover Cluster Help checklists to be sure that all steps and requirements have been met.

If you cannot resolve problems after reading the report, contact the vendor for configuration guidance, or software version compatibility changes. In some cases, a vendor may recommend a specific driver or firmware version that is known to be stable in a clustered environment.

For more information, see "Failover Cluster Help: Validating a Failover Cluster Configuration"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Running the Validate a Configuration Wizard

Questions:

What step is required before you run the Validate a Configuration Wizard?

What are the benefits of running the Validate a Configuration Wizard?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 4 Installing a Failover Cluster on Windows Server 2008 Server Core

Using a Server Core Windows Server 2008 Enterprise or Windows Server 2008 Datacenter installation to create failover clusters provides specific benefits over a full server installation. This lesson describes the benefits of installing a failover cluster on Server Core, and the general process for completing an installation.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Benefits of Installing Failover Clusters on Server Core

Key Points Installing a failover cluster on Server Core results in several key benefits, due to the fact that Server Core only installs essential components. These benefits include:

• A reduced vulnerability attack surface, and a reduced number of required updates.

• Fewer components to manage, and less required disk space for the installation.

• Lower resource overhead, which provides more resources for the clustered instances.

For more information, see "Server Core Installation Option of Windows Server 2008 Step-by-Step Guide"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Supported Roles for Failover Clusters on Server Core

Key Points Because of the reduced overhead and functionality of Server Core, you can cluster only specific roles and features.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Process of Installing a Failover Cluster on Server Core

Key Points Installing a Server Core failover cluster is similar to installing an operating system; however, you must administer Server Core from a command line or from a remote server.

To install a failover cluster on Server Core, you must:

1. Install the operating system.

2. Use netsh to configure the network and Windows firewall settings.

3. Use netdom to join the servers to the domain.

4. Use Diskpart to configure the disk partitions from the command line.

5. Use ocsetup to install the Failover Clustering feature from the command line of each node in the cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


6. Use a server with a full installation of Windows Server 2008 to run the Validate a Configuration Wizard before creating the cluster.

7. Create the cluster from the same server on which the Validate a Configuration Wizard is run, or from the command line of one of the Server Core computers.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Installing a Failover Cluster on Server Core

Questions:

When you install the Failover Clustering feature from the command line, what output is produced?

Why must you use the command line to manage a Server Core computer?


For more information, see "Failover Cluster Help: Understanding Requirement for Failover Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Preparing for a Cluster Installation

Lab Objectives • Install the Failover Clustering feature.

• Validate the failover cluster.

Scenario Fourth Coffee is deploying a new file server cluster. The two servers you will use in the file server cluster have an operating system installed, the clustered disk presented, and network adapters installed. You must prepare the servers for clustering the file server role. Your task is to install the Failover Clustering feature, and complete the Validate a Configuration Wizard.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Installing the Failover Clustering Feature The main tasks for this exercise are to:

• Start the virtual servers.

• Install and verify the Failover Clustering feature.

Task 1: Start the virtual servers 1. Open the 6423A Lab Launcher.

2. Start the 6423A-VAN-DC1 virtual machine.

3. Start the 6423A-VAN-SRV3A virtual machine.

4. Start the 6423A-VAN-SRV3B virtual machine.

Task 2: Install and verify the Failover Clustering feature 1. Log on to 6423A-VAN-SRV3A as Administrator using the password

Pa$$w0rd.

2. Install the Failover Clustering feature.

3. Log on to 6423A-VAN-SRV3B as Administrator using the password Pa$$w0rd.

4. Install the Failover Clustering feature.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Validating the Failover Cluster The main task for this exercise is to validate the failover cluster.

Task 1: Validate the failover cluster 1. Run the Validate a Configuration Wizard.

2. Review the validation report.

3. Record any warnings and errors.

Task 2: Close all virtual machines and discard undo disks 1. For each running virtual machine, close the Virtual Machine Remote Control

window.

2. In the Close box, select Turn off machine and discard changes, and then click OK.

3. Close the 6423A Lab Launcher.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Overview of Failover Cluster Storage Requirements 4-1

Module 4 Overview of Failover Cluster Storage Requirements

Contents: Lesson 1: Overview of Storage Technologies 4-3

Lesson 2: Introduction to Storage Area Networks 4-22

Lesson 3: Planning a Storage Solution for Failover Clusters 4-37

Lesson 4: Configuring an iSCSI Storage Connection 4-43

Lab: Identifying SAN Components 4-49

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

Planning for shared storage is one of the most important aspects of cluster deployment. Lack of proper storage planning can result in unstable clusters.

This module defines basic storage terminology, types of available configurations, and information about Storage Area Networks (SAN).

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Storage Technologies

This lesson describes disk and connectivity types, and provides information on Storage Area Networks (SAN), how SANs can assist with data resiliency and replication, and the differences between SANs and Network Attached Storage (NAS).

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Supported Failover Cluster Storage Types

Key Points Windows Server®°2008 supports only Fibre Channel, Internet SCSI (iSCSI), and Serial Attached SCSI (SAS) storage for failover clusters. Windows Server 2008 does not support Parallel SCSI for clustered disks in two-node clusters, as it did in previous versions of Windows Server. If you create a Windows Server 2008 cluster from hardware that you used with a legacy version of Windows Server, it is important that you verify that Windows Server 2008 supports the storage hardware.

For more information, see "Failover Clustering Help: Validating a Failover Cluster Configuration: Understanding Requirements for Failover Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Storage Type Terminology

Key Points Windows Server 2008 supports three types storage technology for failover clusters:

• Fibre Channel: Transmits SCSI commands over twisted-pair copper wire, or more commonly fiber optic cables. Fibre Channel is commonly found in speeds of 1, 2, or 4 gigabits (GB) per second. Fibre Channel requires specialized interface cards, switches, and cabling to create the network.

• iSCSI: Transmits SCSI commands over Internet Protocol (IP). iSCSI relies on standard Ethernet networking, and requires no specialized hardware. Each device on an iSCSI SAN is given an IP address to communicate with other devices on the iSCSI SAN. Although the iSCSI protocol has more overhead than Fibre Channel, Ethernet networking speeds are now becoming affordable at 10 GB.

• Serial Attached SCSI: Transmits SCSI commands serially in order to provide throughput (up to 6 GB per second,) that is higher than the previous parallel SCSI standard. Serial Attached SCSI is not backward compatible with parallel SCSI equipment.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Clustering"

For more information, see "Storage Technology"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is a Storage Area Network? (SAN)

Key Points A SAN is a set of interconnected devices, such as disks, tapes, and servers, that are connected to a common communication and data transfer infrastructure, such as Fibre Channel. The common communication and data transfer mechanism for a given deployment is called the storage fabric.

The SAN enables multiple servers to access a pool of storage in which any server can potentially access any storage unit. In this environment, management (which determines who is authorized to access which devices,) and sequencing or serialization guarantees, (which determine who can access which devices at what point in time,) play a large role in providing security guarantees.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


A SAN is a network like any other network, such as a local area network (LAN). Therefore, you can use a SAN to connect many different devices and hosts in order to provide access to any device from anywhere. Direct attached storage (DAS) technologies, such as SCSI, are tuned to the specific requirements of connecting mass storage devices to host computers. In particular, they are low-latency, high-bandwidth connections with extremely high data integrity semantics. Conversely, network technology is tuned more to providing application-to-application connectivity in increasingly complex and large-scale environments. Typical network infrastructures have high connectivity, can route data across many independent network segments—potentially over very large distances—and have many network management and troubleshooting tools.

A SAN environment provides the following benefits:

• Centralization of storage into a single pool, which enables storage resources and server resources to grow independently. It also enables storage to be dynamically assigned from the pool when it is required. Storage on a given server can be increased or decreased as needed without complex reconfiguring or re-cabling of devices.

• Common infrastructure for attaching storage, which enables a single common management model for configuration and deployment.

• Storage devices are inherently shared by multiple systems.

• Data can be transferred directly from device to device without server intervention.

• Enables clusters to be built where shared access to a data set is required.

For more information, see "Simple SAN"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Hardware-Based Data Snapshots

Key Points Storage Area Networks can provide advanced features such as data replication and hardware-based snapshots. Hardware-based snapshots enable you to create copies of the failover cluster data, which is transparent to the clustered applications.

Clustered applications, such as database and messaging servers, often store a large amount of data on the clustered disk. These services not only require backups, but also require effective restores in the event of data loss.

Although traditional tape-based backups continue to improve in performance, hardware-based snapshots might be a more effective solution. In the event of data failure, you can attach the snapshot of the data to the clustered servers in a matter of minutes to return an application to service.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Data replication is also a popular solution for SANs. A replication is similar to a snapshot; however, replication makes a copy of data across two SANs or between two storage systems. You can use replication to facilitate disaster recovery scenarios with multi-site failover clusters.

Hardware-based snapshot and replication methodologies and processes can vary between hardware vendors; however, the overall concepts stay the same.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How a SAN/DAS Provides Data Redundancy

Key Points Data redundancy is important in a failover cluster where the goal for the solution is to increase availability. You can use a variety of methods for improving storage systems performance and reliability. The following is a list of methods for writing data to storage:

RAID Type Description

RAID 0 Data is striped across multiple disks. Transactional performance is very good because multiple disks can respond to read requests. No parity information needs to be calculated; therefore, disk writes are efficiently handled. Because there is no redundancy for any of the data stored on the RAID set, when a disk fails in a RAID 0 array, all data is lost.

RAID 1 Data is mirrored across two disks. Transactional performance is very good because either disk in the mirror can respond to read requests. No parity information needs to be calculated; therefore, disk writes are efficiently handled. Each disk in the mirrored set must perform the same write.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

RAID Type Description

RAID 10 Data is striped (RAID 0) across mirrored (RAID 1) sets. RAID 0+1 is not the same as RAID 10. Transactional performance with RAID 10 is very good because either disk in the mirror can respond to read requests. No parity information needs to be calculated; therefore, disk writes are efficiently handled. Each disk in the mirrored set must perform the same write.

When a disk fails in a RAID 10 array, write-performance is not affected because there is still a member of the mirror that can accept writes. In this situation, reads are moderately affected because only one physical disk can respond to read requests. When the failed disk is replaced, the mirror is again established, and the data must be copied or rebuilt.

RAID 5 Calculates parity that surviving member data can use to re-create the data on a failed disk. Writing to a RAID 5 array causes up to four I/Os for each I/O to be written, and the parity calculation can consume controller or server resources. Transactional performance with RAID 5 can still be good, particularly when using a storage controller to calculate the parity.

When a disk fails in a RAID 5 array, the array is in a degraded state, performance is decreased, and latencies are higher. This occurs because most arrays spread the parity information equally across all disks in the array, and it can be combined with surviving data blocks to reconstruct data in real time. Both reads and writes must access multiple physical disks to reconstruct data on a lost disk; during a failure on a RAID 5 array, this increases latency and reduces performance. When the failed disk is replaced, the parity and surviving blocks are used to reconstruct the lost data. This is a lengthy process that can take a number of hours to complete.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is the Difference Between a SAN and a NAS?

Key Points A SAN and a NAS differ in several key areas: a SAN is a block-level access storage solution that uses Fibre Channel, iSCSI, or SAS, a NAS solution uses file-level access such as Common Internet File System (CIFS), File Transfer Protocol (FTP), network file system (NFS), and Hypertext Transfer Protocol (HTTP). Windows Server 2008 supports SAN in a failover cluster, yet does not support NAS for failover clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The following table summarizes the difference between the two technologies:

Parameter SAN NAS

Application Server Access Methods

Block-level access File-level access

Communication Protocol

SCSI over Fibre Channel iSCSI, which is SCSI over IP and Serially Attached SCSI (SAS)

CIFS, NFS, FTP, and HTTP

Network Physical Technology

Typically storage-specific, such as Fibre Channel, but may also be high-speed Ethernet

General purpose LAN, such as Gigabit Ethernet

Support for Windows Server 2008 Failover Clustering

Several types of SAN topologies are supported for clustered disks

Not supported for failover clusters

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Managing Clustered Disks in Windows Server 2008

Key Points Windows Server 2008 clusters offer two main components to disk and path management: Virtual Disk Service (VDS), and Microsoft Multipath I/O (MPIO).

VDS is a set of application programming interfaces (APIs) that provide a single interface for managing disks. This provides an end-to-end solution for managing storage hardware and disks, and for creating volumes on those disks.

VDS uses two sets of providers to manage storage devices. The built-in VDS software providers enable you to manage disks and volumes at the operating system level. Hardware providers that the hardware vendor supplies enable you to manage hardware RAID arrays. Each hardware vendor writes a VDS hardware provider that translates the general-purpose VDS APIs into specific instructions for their hardware. VDS simplifies storage management, because storage applications no longer need to account for the specific hardware that is being targeted. Because of VDS, you no longer need to use different storage applications to manage each different storage device.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


MPIO provides redundant failover and load-balancing support for multipath disk devices. These devices may be disks or logical units.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Multipathing Works

Key Points In a multipath environment, you connect two host bus adapters (HBA) or network interface cards (NIC) to two separate fabrics or networks, so that the storage is still accessible if there is a failure. MPIO provides multipathing and aggregation of paths to remote disks.

MPIO storage driver provides:

• Dynamic configuration and replacement of devices: The operating system must be able to dynamically discover and configure adapters that are connected to the same storage media, in order to support multiple paths to the same storage device.

• Generic device specific module: Microsoft supplies a generic device specific module (DSM) that interacts with the multipath bus driver on behalf of the storage device.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Dynamic load balancing: The multipath software supports the ability to distribute I/O transactions across multiple adapters. The DSM is responsible for load-balancing policy for its storage device.

• Fault tolerance: Multipath software can function in a fault-tolerant mode in which only a single channel is active.

For more information, see "Microsoft Storage Technologies: Multipath I/O"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Comparing SAN Disk Storage Connectivity Types

Key Points SAN disk storage can be different than the type of connectivity that is available from the storage to the host servers. SAN disk storage types include:

• Fibre Channel disks with dual port Fibre Channel connections to provide connectivity redundancy; however, these are more expensive than other types of disk-based storage. Fibre Channel disks are designed for high-duty cycle, and a high mean time to failure (MTTF).

• Fiber Channel ATA (FATA) disks that are based on Serial Advanced Technology Attachment (SATA) standards; however, they are connected via Fibre Channel. These drives typically have a much lower MTTF rating than comparably sized Fibre Channel or SAS-based disks, and are primarily for archival data.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Serial ATA (SATA) disks, that have the same purpose as FATA-based disks; however, they are less expensive and connect via a SATA bus.

• Serial Attached SCSI (SAS) disks that are high duty cycle disks with a higher MTTF than FATA and SATA; they are often less expensive than comparably sized Fibre Channel disks.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Comparing SAN Connectivity Types

Key Points Windows Server 2008 supports three SAN connectivity types for failover clusters:

• Fibre Channel connectivity requires specialized host adapters, cables, and switches to provide a specialized, fast interconnect between servers and storage. Managing Fibre Channel hardware may require training because of differences with other SAN technologies.

• iSCSI connectivity is less expensive than Fibre Channel and requires no specialized hardware. Because iSCSI wraps SCSI commands in TCP/IP packets and transmits over standard Ethernet, you can expect higher latency.

• SAS connectivity is inexpensive; however, it is more limited in the number of hosts that you can connect, and less flexible in how you can configure failover cluster support.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Introduction to Storage Area Networks

Storage Area Networks (SAN) are the basis for most Windows Server 2008 failover cluster storage solutions. This lesson provides details of SAN solutions, and the components for Fibre Channel and iSCSI solutions.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


SAN Terminology

Key Points Use specific terminology when discussing SANs.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How a Storage Controller Works

Key Points A storage controller is a device that houses a set of disks, and provides a single connection (which is redundant and highly available) to a SAN fabric.

Modern controllers usually provide some level of redundancy for data. For example, many controllers offer a wide variety of RAID levels, such as RAID 1, RAID 5, and RAID 0+1 to ensure data availability in the event of the failure of an individual disk drive. The controller presents a virtual view of highly available storage devices, called logical devices or logical unit numbers (LUN), to the hosts. A controller requires at least two ports for highly available storage controllers.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Fibre Channel SAN Terminology

Key Points Use specific terms when discussing Fibre Channel configurations.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Switched Fibre Channel Configuration

Key Points In a Fibre Channel switched fabric (FC-SW), Fibre Channel switches connect devices in a many-to-many design. When a host or device communicates with another host or device, the source and target create a point-to-point connection between them, and then communicate directly with each other. The fabric itself routes data from the source to the target. In an FC-SW, the media is not shared. Any device can communicate with any other device, assuming it is not busy, and communication occurs at full bus speed regardless of whether other devices and hosts are communicating.

When a host or device is powered on, it must first log in to the fabric. This enables the device to determine the type of fabric (a fabric supports a specific set of characteristics,) and provides a fabric address to a host or device. A given host or device continues to use the same fabric address while it is logged on to the fabric, and the fabric address is guaranteed to be unique for that fabric. In order for a host or device to communicate with another device, it must establish a connection to that device before transmitting data. The switches route the packets in the fabric.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Enterprise Design for SAN"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Highly Available Fibre Channel Configurations: Mesh

Key Points In a full mesh design, each switch connects directly to every other switch. Full-mesh design provides any-to-any connectivity, distributes traffic evenly, and consumes low overall bandwidth.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The following table provides pros and cons for a mesh design.

Pros/Cons Description

Pros

• Highly scalable and available SAN configuration.

• Management is simplified; the configuration is a highly available single fabric. Therefore, you must manage only one set of zoning information and one set of security information.

• The fabric itself can route around failures, such as link failures and switch failures.

Cons

• Hosts with multiple adapters must run additional multipathing software to ensure that the host gets a single view of the devices where there are multiple paths from the HBAs to the devices.

• Management errors are propagated to the entire fabric.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Highly Available Fibre Channel Configurations: Core-to-Edge

Key Points The core-to-edge SAN design enables the SAN fabric to scale. The dual fabric design provides fault tolerance by eliminating single points of failures. From a logical perspective, the core-to-edge SAN fabric is comprised of two layers of fiber switches: the edge and the core. Each fabric has edge switches that connect to server hosts, and core switches that connect to the storage subsystems.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The following table provides pros and cons for the core-to-edge SAN design.

Pros/Cons Description

Pros

• Resilient to management or user errors. For example, if security is changed or zones are deleted, the configuration on the alternate fabric is untouched and can be re-applied to the broken fabric.

Cons

• Managing multiple independent fabrics can be costly and error prone. Each fabric should have the same zoning and security information to ensure a consistent view of the fabric regardless of the communication port chosen.

• Hosts and devices must have multiple adapters. In the case of a host, multiple adapters are typically treated as different storage buses. Additional multipathing software, such as MPIO, is required to ensure that the host receives a single view of the devices across the two HBAs.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Zoning with Fibre Channel Switches

Key Points You use zoning to control which device Worldwide Names (WWNs) can see other devices in a SAN. There are a number of server, storage, and switch approaches to zoning.

On any server, there are various mechanisms that control what devices an application can see, and whether or not the application can talk to another device. At the lowest level, the HBA firmware or driver has a masking capability to control whether the server can see other devices. In addition, the operating system can control which devices it tries to mount as a storage volume. You can also use third-party software for volume management, clustering, and file-system sharing, which can also control application access.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Most storage controllers have a form of selective presentation for storage zoning. You configure the controller with lists of which servers can access which LUNs on which ports; the controller ignores or rejects access requests from devices that are not in these lists.

Fibre Channel switch-zoning controls which devices on which ports can access other devices or ports.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


iSCSI SAN Terminology

Key Points You use specific terms to describe iSCSI connectivity, clients, and configuration changes.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


iSCSI Storage Discovery

Key Points Finding available storage controllers and LUNs can be difficult on a large iSCSI network; however, there are provisions to assist in locating storage controllers.

• SendTargets retrieves a list of available targets from the target device. It requires that you to know the IP address of the storage device that is hosting the targets, and is only suitable in smaller iSCSI environments.

• Internet Storage Name Service (iSNS), which is a Windows Server 2008 feature similar to DNS, enables you to locate a target on a number of target devices. iSNS contains three distinct services:

• Name Registration Service: Enables initiators and targets to register and query the iSNS server directory for information about initiator and target ID and addresses.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Network Zoning and Logon Control Service: Allows you to restrict iSNS initiators to zones, so that they cannot discover target devices outside their discovery domains. This prevents initiators from accessing storage devices that are not intended for their use. Logon control allows targets to determine which initiators can access them.

• State Change Notification Service: Enables iSNS to notify clients of changes in the network, such as the addition or removal of targets, or changes in zoning membership. Only initiators that you register to receive notifications will get these packets, which reduces random broadcast traffic on the network.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Planning a Storage Solution for Failover Clusters

This lesson outlines recommendations for Windows Server 2008 failover cluster SAN performance and availability.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Planning a Failover Cluster SAN

Key Points To plan a failover cluster SAN that meets business needs, you must document the availability and performance goals for the applications that you will cluster. Availability and performance goals include application performance requirements, and business requirements for availability of those applications.

After you determine the performance and business requirements, you identify the hardware that meets the requirements. Make sure that each hardware component you choose has the “Certified for Windows Server 2008” logo.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Considerations for SAN Performance and Availability


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Recommendations for SAN Performance

Key Points Because recommendations for SAN performance can vary by vendor, we recommend that you consult the vendor’s documentation, where appropriate. To provide the appropriate performance for the clustered applications, review some of the following factors that affect all storage systems:

• Verify that adequate I/O is available

• Some SAN vendors enable you to add storage controllers to increase the amount of available throughput.

• Some SAN vendors enable you to add cache memory to the storage controllers, which allows data to be written or retrieved from memory instead of waiting for the hard disks to respond.

• Because all data gets written to disks eventually, be sure that you have enough disks to sustain the data throughput so that performance does not suffer.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Verify hardware configuration

• The storage adapters in each node should have the same firmware and driver versions.

• Follow the storage vendor’s recommendations for firmware and driver versions.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Recommendations for SAN Availability

Key Points If the underlying storage does not provide the availability that you require for the failover cluster, the cluster availability goals will not be met. To provide the appropriate availability for the clustered applications, consider some of the common factors that affect all storage systems:

• Ensure redundant components: When available and needed to meet availability requirements, you should consider redundant controllers, memory, and power supplies. This also extends to the Fibre Channel or Ethernet switches. The hosts should have multiple HBA or network adapters to provide multipath connectivity to the storage.

• Ensure resilient configurations: Be sure that you configure the LUNs with the appropriate level of redundancy, and the appropriate number of spare disks to take over in the event of a failure.

• Ensure parallel configurations: Be sure that each host has the same SAN interface adapters with the same firmware, driver, and DSM versions.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 4 Configuring an iSCSI Storage Connection

In some environments, you might choose iSCSI as the SAN storage solution due to its low cost of entry, and adequate performance. This lesson describes the iSCSI components on a Windows Server 2008 computer, and explains how to configure them.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Windows Server 2008 iSCSI Components

Key Points Windows Server 2008 includes three iSCSI components:

iSCSI Component Description

iSCSI initiator Can either be hardware or software-based, and provides a client to the iSCSI storage. Hardware-based iSCSI initiators are typically iSCSI HBA. Windows Server 2008 includes a software-based initiator to connect to an iSCSI SAN without any additional hardware or software.

MPIO DSM A generic set of device-specific modules for MPIO that works with a variety of storage vendors, and is shipped with Windows Server 2008.

Microsoft Internet Storage Name Service

A Windows Server 2008 feature, similar to DNS, that allows a client to locate a target on a number of target devices.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


iSCSI Targets

Key Points You can find iSCSI targets with hardware, third-party software, or natively on a Windows Storage Server edition hardware appliance.

• Many storage vendors provide iSCSI targets as part of their hardware solution.

• Many solutions provide iSCSI targets by loading third-party software on an existing operating system.

• Windows Storage Server appliance provides a hardware appliance with iSCSI target capabilities.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Persistent Bindings and Reservations

Key Points You use persistent bindings to configure iSCSI targets on hosts. A persistent binding maintains a connection for cluster nodes to a target, between reboots.

You use persistent reservations to provide contention resolution when multiple initiators have access to the same LUN. Each initiator has a specific reservation key. When the initiator takes ownership of the LUN, it registers the assigned key to block all other access to the LUN. The reservation keys are automatically assigned.

For more information, see "SPC-3 Letting persistent reservations ignore initiator ports"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Configuring the iSNS Feature

Question:

In what scenarios might you decide to install an iSNS server?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Configuring the iSCSI Software

Questions:

Why might it be necessary to choose persistency when creating bindings?

When might choosing MPIO on the binding not have any affect?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Identifying SAN Components

Lab Objectives • Identify Fibre Channel Storage Area Network (SAN) components.

• Configuring Internet SCSI (iSCSI) storage connections.

Scenario Use the following scenarios to complete the lab exercises.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Identifying Fibre Channel Storage Area Network (SAN) Components The main task for this exercise is to identify components and points of failure based on the given scenarios.

Scenario 1

Tailspin Toys is designing a simple Fibre Channel-based storage area network that will initially include a file server cluster. You are given the preliminary diagram of the new SAN, and have been asked to identify where the Host Bus Adaptors (HBAs), switches, storage controller, and failure points might be.

1. Identify each component in the diagram using the word list shown.

2. Identify any single points of failure.

Host

Host Bus Adapter

Storage Controller

Fibre Channel Switch

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Scenario 2

Tailspin Toys is designing a simple iSCSI-based storage area network that will initially include a file server cluster. You are given the preliminary diagram of the new SAN, and have been asked to identify where the network adapters, switches, storage controller, and failure points might be.

1. Identify each component in the diagram using the word list shown.

2. Identify any single points of failure.

•

Host

Gigabit Ethernet Switch

Storage Controller

Network Adapter

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Configuring Internet SCSI (iSCSI) Storage Connections The main tasks for this exercise are to:

• Start the virtual machines.

• Add the iSCSI target portal to 6423A-VAN-SRV3A.

• Add a persistent binding on 6423A-VAN-SRV3A.

• Add the iSCSI target portal to 6423A-VAN-SRV3B.

• Add a persistent binding on 6423A-VAN-SRV3B.

Task 1: Start the virtual machines 1. Open the 6423A Lab Launcher.




Task 2: Add the iSCSI target portal to VAN-SRV3A 1. Log on to VAN-SRV3A as Administrator using the password Pa$$w0rd.

2. Add the iSCSI target portal 192.168.10.12 to VAN-SRV3A.

Task 3: Add a persistent binding on VAN-SRV3A • Add a persistent binding for iqn. 1991-05.com.microsoft:van-dc1-van-srv3a-

target on VAN-SRV3A.

Task 4: Add the iSCSI target portal to VAN-SRV3B 1. Log on to VAN-SRV3B as Administrator using the password Pa$$w0rd.

2. Add the iSCSI target portal 192.168.10.12 to VAN-SRV3B.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 5: Add a persistent binding on VAN-SRV3B • Add a persistent binding for iqn. 1991-05.com.microsoft:van-dc1-van-srv3b-

target on VAN-SRV3B.


window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Configuring a Failover Cluster 5-1

Module 5 Configuring a Failover Cluster

Contents: Lesson 1: Creating a New Failover Cluster 5-3

Lesson 2: Managing a Failover Cluster 5-7

Lesson 3: Verifying Failover Functionality 5-14

Lab: Creating and Administering a Cluster 5-21

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

After you verify the cluster hardware and software by using the Validate a Configuration Wizard, you are ready to create the cluster, cluster resources, and clustered applications.

This module describes how to create a cluster, and how to perform basic management tasks for the newly created cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Creating a New Failover Cluster

This lesson describes how to create a cluster and verify cluster functionality.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Creating a Cluster Resource

Key Points After you complete the Validate a Configuration Wizard, you create a cluster resource. To create the cluster resource, you must log on to a server with a domain account that has local administrator permissions on each of the cluster nodes.

Start the Create a Cluster Wizard from the Failover Cluster Management Microsoft® Management Console (MMC). The wizard prompts you for a list of server nodes that will be members of the cluster. After entering each of the nodes, provide a cluster name and IP addresses.

For more information, see "Failover Cluster Help: Add Resource Type Dialog Box"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Creating a Cluster

Question:

Is there a reason that you might run the Create a Cluster Wizard from a server that is not part of the cluster?

For more information, see "Windows Server 2008 Failover Cluster Step-by-Step"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Verifying Cluster Functionality

Questions:

What information should you review in the failover cluster installation report?

Why is it important to examine the installation report closely before you deploy clustered applications?

For more information, see "Windows Server 2008 Failover Cluster Step-by-Step"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Managing a Failover Cluster

Built-in management tools enable you to manage a cluster. This lesson outlines the failover cluster management options that help you maintain and configure the cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Management Tools

Key Points To manage the cluster, you can use either the Failover Cluster Management tool or the cluster.exe command line tool. Typically, you will use the Failover Cluster Management tool; however, in addition to management tasks, you might use the command line tool for in-depth troubleshooting and logging activities.

You must use the cluster.exe tool from a computer running the Server Core installation option of Windows Server 2008. To have the failover cluster management tools available, you must install the Failover Clustering feature, or the Remote Server Administration Tools, on a Windows Server 2008 or a Windows Vista® Service Pack 1 computer.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Configuring Cluster Properties

Key Points You can perform several failover cluster management tasks. These tasks range from adding and removing cluster nodes, to modifying the quorum settings.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Modifying Quorum Settings

Key Points Usually, the Create a Cluster Wizard chooses the quorum mode for your environment based on the failover cluster configuration. However, you might need to modify quorum settings using the Configure Cluster Quorum Wizard, in the Failover Cluster Management tool. If you have an Exchange Server 2007 Continuous Copy Replication cluster configuration, you will need to manually set the quorum mode to Node and File Share Majority.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Managing Cluster Nodes

Key Points There are three aspects to managing cluster nodes:

• You can add a node to an established failover cluster by selecting Add Node from the Failover Cluster Management Actions pane. The Add Node Wizard will prompt you for information about the additional node.

• Pausing a node prevents resources from being failed over or moved to the node. You typically pause a node when a node is undergoing maintenance or troubleshooting.

• Evicting is an irreversible process for a cluster node. Once you evict the node, it must be re-added to the cluster. You evict nodes when a node is damaged beyond repair, or is no longer needed in the cluster. If you evict a damaged node, you can repair or rebuild it, and then add it back to the cluster using the Add Node Wizard.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Help"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Using Quorum Configuration Options

Question:

When might it be appropriate to change the quorum mode?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Verifying Failover Functionality

This lesson describes how to specify when and how failovers occur.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Failover Cluster Scenarios

Key Points You need to recognize situations that trigger failover, so you can act accordingly. For example, if an application failure causes a failover, you must determine the root cause of the application failure before allowing the application to fail back to the original cluster node. In all failure cases, to maintain the failover cluster’s stability, you must repair unstable nodes before they can host resources again.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Preferred Owners

Key Points You can create a preferred owners list in order to control which nodes host clustered applications. Using preferred owners ensures that applications run on server nodes where they will perform the best, and not interfere with other clustered applications.

For example, in an Active/Active cluster with two clustered file servers, one clustered file server will have one cluster node as the preferred owner, and the other clustered file server will have the other node as a preferred owner. This spreads the load across the cluster nodes.

For more information, see "Failover Cluster Help: Modifying Settings for a Clustered Service or Application"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Modifying Failover and Failback Settings

Key Points You might need to adjust the preferred owners and failback settings in order to control how the cluster responds when the application or service fails. You only use these settings when a clustered application or service failure occurs. You change these settings from the clustered application property sheets.

The following table includes sample failover and failback settings.

Settings Effect

General tab, Preferred owner: Node1

Failover tab, Failback setting: Allow failback (Immediately)

If the service or application fails over from Node1 to Node2, when Node1 is again available, the service or application will fail back to Node1.

Failover tab, Maximum failures in the specified period: 4

Failover tab, Period (hours): 6

In a six-hour period, if the application or service fails 10 times or less, it will be restarted or failed over each time. If the application or service fails a fifth time in the six-hour period, it will be left in the failed state.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Failover Cluster Help: Modifying Settings for a Clustered Service or Application"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Performing Manual Failovers

Key Points Failovers are not only caused by hardware or software failures; you might need to perform a manual failover to maintain or troubleshoot nodes. The manual process moves a clustered application from one node to another.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Testing Auto Failovers

Question:

Why would you test automatic failovers on production failover clusters?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Creating and Administering a Cluster

Lab Objectives • Create a cluster

• Manage a failover cluster

Scenario Fabrikam has two servers that provide clustered file services. You will complete the deployment of Fabrikam’s failover cluster. The domain controller is available to act as the cluster’s file share witness.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Creating a Cluster In this exercise, you will start the virtual servers, and then create and verify a failover cluster.

The main tasks for this exercise are to:

• Start the virtual servers.

• Add the iSCSI target portal to VAN-SRV5A.

• Add the iSCSI target portal to VAN-SRV5B.

• Configure the shared disks.

• Validate the failover cluster.

• Use the Create a Cluster Wizard to build a simple failover cluster.

• Verify the successful creation of the cluster.

Task 1: Start the virtual servers 1. Open the 6423A Lab Launcher.




Task 2: Add the iSCSI target portal to VAN-SRV5A 1. Log on to VAN-SRV5A as Administrator using the password Pa$$w0rd.

2. Add the iSCSI target portal 192.168.10.12 to VAN-SRV5A.

3. Add a persistent binding for iqn. 1991-05.com.microsoft:van-dc1-van-srv5a-target on VAN-SRV5A.

Task 3: Add the iSCSI target portal to VAN-SRV5B 1. Log on to VAN-SRV5B as Administrator using the password Pa$$w0rd.

2. Add the iSCSI target portal 192.168.10.12 to VAN-SRV5B.

3. Add a persistent binding for iqn. 1991-05.com.microsoft:van-dc1-van-srv5b-target on VAN-SRV5B.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 4: Configure the shared disks 1. On VAN-SRV5A, in Server Manager, access Disk Management.

2. Bring Disk 1 online, and then initialize the disk.

3. Create a new simple volume on the disk using all the disk space, and label the volume Data.

4. On VAN-SRV5B, in Server Manager, access Disk Management.

5. Bring Disk 1 online.

Task 5: Validate the failover cluster 1. Run the Validate a Configuration Wizard.

2. Review the validation report.

3 Record any warnings and errors.

Task 6: Use the Create a Cluster Wizard to build a simple failover cluster • Use the Create a Cluster Wizard to create a simple failover cluster with the

following information:

• Cluster name: Cluster1

• IP Address: 10.10.0.125

• Subnet Mask: 255.255.255.0

Task 7: Verify the successful creation of the cluster 1. In the /Windows/Cluster/Reports directory, open the file CreateCluster.mht.

2. Verify that The cluster was successfully created appears.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Managing a Failover Cluster In this exercise, you will modify the cluster quorum mode.


• Use the cluster administration tools to manage the failover cluster.

• Pause and resume the server.

Task 1: Use the cluster administration tools to manage the failover cluster • Modify the quorum mode for Cluster1 using the following:

• Quorum mode: Node and File Share Majority

• File Share Majority: \\VAN-DC1\FSW5

Task 2: Pause and resume VAN-SRV5A 1. Pause VAN-SRV5A.

2. Verify that VAN-SRV5A has been paused.

3. Resume VAN-SRV5A.


window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Configuring Cluster Resources and Server Roles 6-1

Module 6 Configuring Cluster Resources and Server Roles

Contents: Lesson 1: Configuring Cluster Resources 6-3

Lesson 2: Implementing Failover Clusters for Server Roles Using Failover Cluster Management 6-9

Lesson 3: Clustering Server Roles Using Windows Server Core 6-15

Lab: Clustering Server Roles and Features 6-22

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

This module describes how to configure cluster resources and server roles using both full server and Server Core installations.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Configuring Cluster Resources

Cluster resources are one of the building blocks of a clustered service or application.

This lesson describes how to customize options for cluster resources that determine how resource failure will affect the cluster, and how resources should attempt to recover automatically.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Cluster Resources?

Key Points In a failover cluster, cluster resources provide functionality to the clustered applications. Resources are the building blocks of a clustered application or a service. When you configure resources together as a clustered application, they can fail between nodes as a group.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Key Windows Server 2008 Resources

Key Points Windows Server®°2008 provides resources that you can use to cluster functions and features. For example, you can cluster IP version 6 (IPv6) addresses, file services, and virtual machines. Applications such as Microsoft Exchange Server and Microsoft SQL Server® add special resources that support the operation of the application.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Resource Dependencies?

Key Points Resource dependencies perform the following:

• Determine the order in which resources can start and stop. For example, if the network name resource is dependent on an IP address resource, the network name must stop before the IP address stops.

• Define bindings. For example, if a network name resource is dependent on an IP address resource, then that address will be registered in both the Domain Name System (DNS), and Windows Internet Name Service (WINS), with that associated network name resource.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How Resource Failover Works for Specific Resources

Key Points During failover, each of the resources is taken offline in the order set by the dependencies. After all resources are offline, the cluster moves the entire group ownership from the active node to the passive node. After the group ownership changes, the resources are started in the reverse order, which ensures that dependent resources are started first. The length of time for a failover depends on the cluster node’s transaction load, and the resources that are being failed-over.

For more information, see "Failover and Failback"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Configuring Cluster Resource Policies

Key Points You can customize a resource’s response to a failover. For example, if a disk resource fails on the active node, you might want the entire group of resources to fail over to another node. Or you might want to customize the resource policies to attempt to restart the resource several times before failing the service or application over to another cluster node.

For more information, see "Failover Clustering Help: User Interface: The Failover Cluster Management Snap-In <Resource> Properties: Policies Tab"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Implementing Failover Clusters for Server Roles Using Failover Cluster Management

Resources are one part of a clustered service or application. You group resources to provide a service or application on a failover cluster.

This lesson describes how to implement a clustered application.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Cluster Resources and Server Roles

Key Points A clustered application is a group of resources that provide a service. A file server, a database server that provides a service, and an application, are all examples of clustered applications.

Before you cluster a Windows Server 2008 role, you use Server Manager to install the role on each cluster node. Upon installation, these roles add the resources that are required to support that role.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Process to Cluster Server Roles

Key Points To cluster an application or service, you first create a cluster, and on each node, install the roles that will be clustered. Then, using the Failover Cluster Management snap-in, you cluster the role. After you cluster the application, you can configure the failover policy and the application.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Roles That Benefit from Clustering


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Common Clustered Server Roles

Key Points Although you can cluster a number of roles and features, certain roles are more commonly clustered. These roles are more common because they are critical to business operations. For example, a file server can store data critical to a specific business process; therefore, it is an excellent candidate for clustering. You can also cluster other applications that are not server roles, such as e-mail and database services, in order to provide high availability.

For more information, see "Failover Clustering Help: Configuring a Service or Application for High Availability"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Clustering Print Services

Question:

What corporate scenarios might benefit from clustering print services?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Clustering Server Roles Using Windows Server Core

Server Core provides a lower overhead environment for clustering. This leaves more resources available, and requires fewer updates to the nodes. This lesson describes how to implement a Server Core failover cluster, which is similar to a full installation, but has limitations and uses different methods for the configuration.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Benefits of Server Core Clustering

Key Points Server Core requires less maintenance than a full installation, because it installs fewer components. It also has a reduced attack surface, which allows fewer components to be exploited, and requires fewer updates if vulnerabilities are discovered. Server Core also requires less disk space and memory, which frees more resources for the clustered applications and services.

For more information, see "Windows Server 2008 Product Overview"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Managing Windows Server Core

Key Points You can use the following command-line tools to manage Server Core:

Tool Description

OcSetup.exe Installs or removes optional Windows components.

Cluster.exe Creates and configures cluster settings.

iSCSIcli.exe Manages Internet SCSI (iSCSI) client configurations.

OcList.exe Lists the server roles and optional features that are available for use with Ocsetup.exe.

Netsh.exe Modifies network settings.

Diskpart.exe Manages disks.

Sc.exe Provides status and manages services.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Additionally, you can run the following standard management tools from a remote server with a full installation:

Tool Description

Failover Cluster Management snap-in Modifies and manages failover cluster settings.

Group Policy Management snap-in Manages Active Directory group policy objects.

Computer Management snap-in Manages computer settings and services.

For more information, see "Windows Server 2008 Step-by-Step Guides"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Server Core Clustered Roles

Key Points A number of roles and functions can benefit from Server Core. One of the main reasons to use Server Core is the Hyper-V feature. With its reduced memory, CPU, and hard-drive resource overhead, Server Core can provide additional resources to run the virtual machines. Additionally, other roles such as Print Services, File Services, DHCP Server, and DNS Server can also benefit from running on Server Core.

For more information, see "Server Core Installation of Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Installing the Print Services Role Using ocSetup.exe

Question:

In what other instances might you want to install roles using ocSetup.exe?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Clustering the File Services Role on Windows Server Core

Question:

Installing the Failover Clustering feature, creating the cluster, and clustering the file services are some of the steps you need to take to create a failover cluster on Server Core. What other steps do you need to take when you create a failover cluster on Server Core?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Clustering Server Roles and Features

Lab Objectives • Cluster the Print Services role using Failover Cluster Management.

• Configure cluster resources.

• Cluster the File Services role on Windows Server Core.

Scenario Use the scenarios on the following pages to complete the lab exercises.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Clustering the Print Services Role Using Failover Cluster Management Fabrikam has two servers configured as a failover cluster. You have to cluster the Print Services role on these two servers. The Print Services role already is installed on VAN-SRV6A and VAN-SRV6B.


• Start the virtual machines, and then log on.

• Configure the printer disk.

• Cluster the Print Services role.

• Failover the VAN-Print clustered service from VAN-SRV6A to VAN-SRV6B.

Task 1: Start the virtual machines, and then log on 1. On your host machine, click Start, point to All Programs, point to Microsoft

Learning, and then click 6423A. The Lab Launcher starts.

2. In the Lab Launcher, next to 6423A-VAN-DC1, click Launch.

3. In the Lab Launcher, next to 6423A-VAN-SRV6A, click Launch.

4. In the Lab Launcher, next to 6423A-VAN-SRV6B, click Launch.

5. Log on to VAN-SRV6A as Administrator with the password Pa$$w0rd.

6. Log on to VAN-SRV6B as Administrator with the password Pa$$w0rd.

7. Minimize the Lab Launcher window.

Task 2: Configure the printer disk 1. On VAN-SRV6A, in Server Manager, start Disk Management.

2. Bring Disk 2 online, and then initialize the disk.

3. Create a new simple volume on the disk using all the disk space, and then label the volume Printer1.

4. On VAN-SRV6B, in Server Manager, start Disk Management.

5. Bring Disk 2 online.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 3: Cluster the Print Services role 1. Click Start, click Administrative Tools, and then click Failover Cluster

Management. (If the User Account Control dialog box appears, confirm that the correct action displays, and then click Continue.)

2. In the console tree, select VAN-CLUSTER01, and then select Storage.

3. In the Actions pane, click Add a disk, and then add Disk 2.

4. In the VAN-CLUSTER01 cluster, create a print server clustered application with the following settings:

• Name: VAN-Print

• IP Address: 10.10.0.108

• Storage volume: Cluster Disk 2

Task 4: Failover the VAN-Print clustered service from VAN-SRV6A to VAN-SRV6B 1. In the console tree, select VAN-Print.

2. Move VAN-Print to the passive node.

3. Verify that VAN-Print now shows the current owner as the new node.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Configuring Cluster Resources Fabrikam has two servers configured as a failover cluster. You have to cluster the Print Services role on these two servers. The Print Services role already is installed on VAN-SRV6A and VAN-SRV6B.


• Change the preferred owner of VAN-Print to VAN-SRV6B.

• Change the failback settings to allow only failback to the preferred node between 01:00 and 04:00.

• Close all virtual machines, and discard undo disks.

Task 1: Change the preferred owner of VAN-Print to VAN-SRV6B • Change the preferred owners of VAN-Print to VAN-SRV6B.

Task 2: Change the failback settings to allow only failback to the preferred node between 01:00 and 04:00 • Set the Failback between properties to 1 in the first box, and 4 in the second

box.

Task 3: Close all virtual machines, and discard undo disks 1. For each virtual machine that is running, close the Virtual Machine Remote

Control window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 3: Clustering the File Services Role on Windows Server Core Fabrikam has two servers with Server Core installed that you must configure as a failover cluster. You have to cluster a file service on these two servers. The Print Services role already is installed on VAN-CORE6Aand VAN-CORE6B. The iSCSI target also has been added to the server that you will use as the file share.



• Configure the iSCSI target mappings and shared drives on VAN-CORE6A.

• Configure the iSCSI target mappings and shared drives on VAN-CORE6B.

• Install the Failover Clustering feature on VAN-CORE6A.

• Install the Failover Clustering feature on VAN-CORE6B.

• Create a cluster.

• Create a file share clustered service.

• Add a shared folder to VAN-FILE01.

• Log on to VAN-DC1, and verify the clustered shared folder.

• Close all virtual machines and discard undo disks.




3. In the Lab Launcher, next to 6423A-VAN-CORE6A, click Launch.

4. In the Lab Launcher, next to 6423A-VAN-CORE6B, click Launch.


6. Log on to VAN-CORE6A as Fabrikam\administrator with the password Pa$$w0rd.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 2: Configure the iSCSI target mappings and shared drives on VAN-CORE6A 1. On VAN-CORE6A, use the sc command to configure the msiSCSI service to

start automatically, and then start the service.

2. Use the iscsicli QAddTargetPortal 192.168.12.10 command to add a new portal.

3. Use the iscsicli listtargets command to identify available IQNs.

4. Use the iscsicli QLoginTarget iqnvalue command to log on to the IQN.

5. Use the iscsicli PersistentLoginTarget iqnvalue T * * * * * * * * * * * * * * * 0 where iqnvalue command to make the logon persistent.

6. Use Diskpart to configure a new primary partition on Disk 1 using the drive letter F. Format the partition with NTFS file system.

Task 3: Configure the iSCSI target mappings and shared drives on VAN-CORE6B 1. On VAN-CORE6B, use the sc command to configure the msiSCSI service to

start automatically, and then start the service.

2. Use the iscsicli QAddTargetPortal 192.168.12.10 command to add a new portal.

3. Use the iscsicli listtargets command to identify available IQNs.

4. Use the iscsicli QLoginTarget iqnvalue command to log on to the IQN.

5. Use the iscsicli PersistentLoginTarget iqnvalue T * * * * * * * * * * * * * * * 0 where iqnvalue command to make the logon persistent.

6. Use Diskpart to bring Disk 1 online, and to clear the read-only attribute.

Task 4: Install the Failover Clustering feature on VAN-CORE6A 1. Use start /w ocsetup FailoverCluster-Core to install the Failover Clustering

feature on VAN-CORE6A.

2. To query the status of the cluster service, at the command prompt, type sc qc clussvc, and then press ENTER.

3. Verify that the sc command returns QueryServiceConfig SUCCESS.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 5: Install the Failover Clustering feature on VAN-CORE6B 1. Use start /w ocsetup FailoverCluster-Core to install the Failover Clustering

feature on VAN-CORE6B.

2. Use the Service Control command to query the status of the cluster service. At the command prompt, type sc qc clussvc, and then press ENTER.

3. Verify that the sc command returns QueryServiceConfig SUCCESS.

Task 6: Create a cluster 1. At the command prompt, type cd %systemroot%\cluster, and then press

ENTER.

2. Type cluster VAN-CLUSTER02 /create /ipaddr: 10.10.0.109/255.255.0.0.

Task 7: Add VAN-CORE6B to the cluster • Type cluster VAN-CLUSTER02 /addnode /node:VAN-CORE6B.

Task 8: Create a file share clustered service 1. Log on to VAN-SRV6A as Administrator with the password Pa$$w0rd.

2. Open Failover Cluster Management, and configure it to manage the VAN-Cluster02.fabrikam.com cluster.

3. Modify the quorum configuration for the cluster to use a Node and File Share Majority mode, and to use \\VAN-DC1\FSW5 as the file share.

4. Add a new disk to the cluster.

5. Create a file server clustered service with the following attributes:

• Name: VAN-FILE01

• IP Address: 10.10.0.110

• Storage volume: Cluster Disk 1

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 9: Add a shared folder to VAN-FILE01 1. In the console tree, ensure that VAN-FILE01 is selected.

2. Add a shared folder location F:\.

3. For the Share name, type Public.

Task 10: Log on to VAN-DC1, and verify the clustered shared folder 1. Log on to VAN-DC1 as Administrator with the password Pa$$w0rd.

2. Click Start, click Run, and then type \\VAN-File01\Public.

3. In Windows Explorer, verify that the user can connect to the share.


window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Maintaining Microsoft Failover Clusters 7-1

Module 7 Maintaining Microsoft Failover Clusters

Contents: Lesson 1: Monitoring Failover Clusters 7-3

Lesson 2: Backing Up and Restoring Failover Clusters 7-11

Lesson 3: Troubleshooting Failover Clusters 7-20

Lab: Maintaining Failover Clusters 7-26

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

After deploying a failover cluster, you will need to maintain the cluster, and troubleshoot when issues arise.

This module provides information on how to monitor cluster functionality, troubleshoot the cluster by reviewing logs to identify and solve problems, and adjust settings.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Monitoring Failover Clusters

Failover clusters are more complex than stand-alone servers. This lesson describes how to use standard cluster monitoring tools such as the Event Viewer, and the Reliability and Performance Monitor snap-in.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Tools for Monitoring Failover Clusters

Key Points A variety of tools are available to help you monitor failover clusters. You can use standard Windows Server®°2008 tools, such as the Event Viewer and the Reliability and Performance Monitor snap-in, to review cluster event logs, and performance metrics. You can also use Cluster.exe and Tracerpt.exe to export data for analysis. Additionally, you can use the MHTML-formatted cluster configuration reports and the Validate a Configuration Wizard to troubleshoot problems with the cluster configuration and hardware changes.

For more information, see "Failover Clustering Help: View Events and Logs for a Failover Cluster"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Event Viewer to Monitor Failover Clusters

Key Points When problems arise in the cluster, use the Event Viewer to view events with a Critical, Error, or Warning severity level.

Additionally, informational level events are logged to the Failover Clustering Operations log, which can be found in the Event Viewer in the Applications and Services Logs\Microsoft\Windows folder. Informational-level events are usually common cluster operations, such as cluster nodes leaving and joining the cluster, or resources going offline or coming online.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Viewing Events for the Entire Cluster

Key Points In previous Windows Server versions, event logs were replicated to each node in the cluster. This simplified cluster troubleshooting, because you could review all event logs on a single cluster node. Windows Server 2008 does not replicate the event logs between nodes; however, the Failover Cluster Management snap-in has a Cluster Events option that enables you to view and filter events across all cluster nodes. This feature is helpful in correlating events across cluster nodes.

The Failover Cluster Management snap-in also provides a Recent Cluster Events option that will query all of the Error and Warning events from all of the cluster nodes in the last 24 hours.

You can access additional logs, such as the Debug and Analytic logs, from the Event Viewer. To display these logs, modify the view in the top menu by selecting Show Analytic and Debug Logs.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Event Tracing

Key Points Windows event tracing is a kernel component that is available early after startup, and late into shutdown. It is designed to allow for fast tracing and delivery of events to trace files and to consumers. Because it is designed to be fast, it allows only basic in-process filtering of events based on event attributes.

The event trace log contains a comprehensive accounting of the failover cluster actions. Depending on how you want to view the data, use either Cluster.exe or Tracerpt.exe to access the information in the event trace log.


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Tracerpt.exe to View the Event Trace Log

Key Points Tracerpt.exe will parse the event trace logs only on the node on which it is run. All of the individual logs are collected in a central location. To transform the XML file into a text file or an HTML file that can be opened in Internet Explorer, you can parse the XML-based file using the Microsoft XSL parsing command-line utility msxsl.exe, and an XSL style sheet.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Cluster.exe to Export the Cluster Log

Key Points Cluster.exe provides a cluster log for the entire cluster, or for a specific node in the cluster in a readable text format. The output from Cluster.exe is more flexible than Tracerpt.exe, because you can export trace logs for all nodes in the cluster and save them in a specified location. You can view the exported log using a text editor. To run successfully, Cluster.exe must be run from an administrator-elevated command prompt.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using the Reliability and Performance Monitor Snap-in

Key Points The Performance and Reliability Monitor snap-in allows you to:

• Trend application performance on each node. To determine how an application is performing, you can view and trend specific information on system resources that are being used on each node.

• Trend application failures and stability on each node. You can pinpoint when application failures occur and correlate the application failures with other events on the node.

• Modify trace log settings. You can start, stop, and adjust trace logs, including their size and location.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Backing Up and Restoring Failover Clusters

This lesson describes how to backup a failover cluster, and how to perform an authoritative or a non-authoritative restore.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Backing Up Failover Clusters

Key Points When you back up the cluster configuration, be aware of the following:

• Before putting a cluster into production, you must test your backup and recovery process.

• If you choose to use Windows Server Backup, (the Windows Server 2008 backup feature), you must first add the feature. You can do this during initial configuration tasks, or by using the Add Features Wizard in Server Manager.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Windows Server Backup to Back Up the Cluster Configuration

Key Points Windows Server Backup is the built-in backup and recovery software for Windows Server 2008. In order to complete a successful backup, be aware of the following:

• For a backup to succeed in a failover cluster, the cluster must be running and must have quorum. In other words, enough nodes must be running and communicating (perhaps with a witness disk or witness file share, depending on the quorum configuration,) that the cluster has achieved quorum.

• You must back up all clustered applications. If you cluster a Microsoft SQL Server® database, you must have a backup plan for the databases and configuration outside of the cluster configuration.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• If application data must be backed up, the disks that you store the data on must be made available to the backup software. You can accomplish this by running the backup software from the cluster node that owns the disk resource, or by running a backup against the clustered resource over the network.

• The cluster service keeps track of which cluster configuration is the most recent, and it replicates that configuration to all cluster nodes. If the cluster has a witness disk, the cluster service also replicates the configuration to the witness disk.

For more information, see "Failover Clustering Help: Understanding Backup and Recovery Basics for a Failover Cluster"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Backing Up the Failover Cluster Configuration

Question:

What option must you use in Windows Server Backup to back up the failover cluster configuration?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using a Non-authoritative Restore

Key Points Use a non-authoritative restore when a single node in the cluster has been damaged or rebuilt, and the rest of the cluster is operating properly.

Perform a non-authoritative restore by restoring the system recovery (system state) information to the damaged node. When you restart that node, it will join the cluster and receive the latest cluster configuration automatically.

For more information, see "Failover Clustering Help: Restoring to a Failover Cluster from Backup"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using an Authoritative Restore

Key Points Use an authoritative restore when the cluster configuration must be rolled back to a previous point in time. For example, you would use an authoritative restore if an administrator accidently removed clustered resources or modified other cluster settings.

Perform the authoritative restore by stopping the cluster resource on each of the nodes, and then performing a system recovery (system state) on a single node using the command-line Windows Server Backup interface. After the restored node restarts the cluster service, the remaining cluster nodes can also start the cluster service.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Performing an Authoritative Restore for a Failover Cluster

Questions:

When would you use an authoritative restore?

When would you use a non-authoritative restore?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Addressing Application Recovery


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Troubleshooting Failover Clusters

This lesson defines concepts to help troubleshoot failover clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Troubleshooting Failover Clusters

Key Points To troubleshoot a failover cluster:

• Use the Validate a Configuration Wizard to highlight configuration issues that might cause cluster problems.

• Review cluster events and trace logs to identify application or hardware issues that might cause an unstable cluster.

• Review hardware events and logs to help pinpoint specific hardware components that might cause an unstable cluster.

• Review storage area network (SAN) components, switches, adapters, and storage controllers to help identify any potential problems.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The Process for Troubleshooting Failover Clusters

Key Points When troubleshooting failover clusters, you must:

• Identify the perceived problem by gathering and documenting the symptoms of the problem.

• Identify the scope of the problem so that you can understand what is being affected by the problem, and what impact that effect is having on the application and the clients.

• Gather information so that you can accurately understand and pinpoint the possible problem. After you identify a list of possible problems, you can prioritize them by probability, or the impact of a repair. If the problem cannot be pinpointed, you should attempt to recreate the problem.

• Create a schedule for repairing the problem. For example, if the problem only affects a small subset of users, you can delay the repair to an off-peak time so that you can schedule downtime.

• Complete and test each repair one at a time so that you can identify the fix.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Troubleshooting Techniques


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Troubleshooting Storage Issues

Key Points To troubleshoot SAN issues, start by checking physical connections and each of the hardware component logs. Additionally, run the Validate a Configuration Wizard to verify that the current cluster configuration is still supportable. When you run the Validate a Configuration Wizard, ensure that the storage tests you select can be run on an online failover cluster. Several of the storage tests cause loss of service on the clustered disk when the tests are run.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Troubleshooting Group and Resource Failures

Key Points To troubleshoot group and resource failures:

• Use the Dependency Viewer in the Failover Cluster Management snap-in to identify dependent resources.

• Check the Event Viewer and trace logs for errors from the dependent resources.

• Determine if the problem only happens on a specific node, or nodes, by trying to recreate the problem on different nodes.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Maintaining Failover Clusters

Lab Objectives • Monitor failover clusters.

• Perform backups on a failover cluster.

• Perform an authoritative restore on a failover cluster.

Scenario Fabrikam has asked you to review cluster event logs and perform a test restore to validate the backup process for their existing file cluster. The Windows Server Backup feature is installed on both cluster nodes.

Note: Complete the steps to prepare for this lab after starting the virtual machines and logging on as described in Exercise 1, Task 1.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


To prepare for this lab:

1. On VAN-SRV6A, click Start, click Administrative Tools, and then click Server Manager.

2. Expand Storage, and then click Disk Management.

3. Right-click Disk 2, and then click Online.

4. Right-click Disk 2, and then click Initialize disk. In the Initialize Disk dialog box, click OK.

5. Right-click the unallocated space beside Disk 2, and then click New Simple Volume.

6. On the Welcome page, click Next.

7. On the Specify Volume Size page, click Next.

8. On the Assign Drive Letter or Path page, click Next.

9. On the Format Partition page, in the Volume Label field, type Data, select the Perform a quick format check box, click Next, and then click Finish.

10. On VAN-SRV6B, open Server Manager.

11. Expand Storage, and then click Disk Management.

12. Right-click Disk 2, and then click Online. In the Disk Management error message, click OK.

13. Right-click Disk 2, and then click Online.

14. On VAN-SRV6A, click Start, click Administrative Tools, and then click Failover Cluster Management.

15. In the console tree, expand VAN-CLUSTER01.Fabrikam.com.

16. Click Storage, in the Actions pane, click Add a disk, and then click OK.

17. Click VAN-CLUSTER01.Fabrikam.com.

18. In the Actions pane, click Configure a Service or Application.

19. Review the text on the first page of the wizard, and then click Next.

20. Click Print Server, and then click Next.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


21. Type VAN-Print for the Name and type 10.10.0.109 as the IP Address in the network specified as 10.10.0.0/16, and then click Next.

22. Select Cluster Disk 2 as the storage volume for the print server, and then click Next.

23. Click Next again, and then click Finish.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Monitoring Failover Clusters In this exercise, you will identify cluster events both during a restart and during a failover, and then you will view file cluster dependencies.



• Identify the cluster events that occurred during the last restart.

• Failover VAN-Print, and identify the events that occur.

• View dependencies on the file cluster.






5. Log on to VAN-SRV6A as Administrator with the password Pa$$w0rd.

6. Log on to VAN-SRV6B as Administrator with the password Pa$$w0rd.


8. Complete the steps in the "To prepare for this lab" section above.

Task 2 : Identify the cluster events that occurred during the last restart 1. On VAN-SRV6A, open the Event Viewer from the Administrative tools.

2. In the Event Viewer, browse to Applications and Services Logs\Windows \FailoverClustering\Operational to view the operational logs.

3. In the list of events, locate and review events with the Event IDs 1061, 1125, and 1131.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 3: Failover VAN-Print, and identify the events that occur 1. Under VAN-CLUSTER01, browse to Services and Applications\VAN-Print,

and then move the application to VAN-SRV6B.

2. In the Event Viewer, browse to Applications and Services Logs\Microsoft \Windows\FailoverClusteringOperational to view the operational logs.

3. In the list of events, locate and review events with the Event IDs 1200, 1201, 1203, and 1204.

4. Close the Event Viewer.

Task 4: View dependencies on the file cluster 1. In the Failover Cluster Management snap-in, browse to VAN-CLUSTER01

\Services and Applications.

2. In Services and Applications, right-click VAN-Print, and, in the Actions pane, click Show Dependency Report.

3. In the Dependency Report, identify the VAN-Print dependencies.

4. Close Internet Explorer.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Performing Backups on a Failover Cluster In this exercise, you will verify functionality of the cluster, and create a backup of the cluster configuration database.


• Verify the cluster functionality.

• Create a backup of the cluster configuration database.

Task 1: Verify the cluster functionality 1. In the Failover Cluster Management snap-in, click VAN-CLUSTER01.

2. Verify that there are no application alerts, and that the cluster has quorum.

Task 2: Create a backup of the cluster configuration database 1. Open Windows Server Backup, and in the Actions pane, select Backup Once.

2. In the Backup Once Wizard, select Different Options, and then click Next.

3. Select Custom, and then click Next.

4. Verify that Local Disk (C:) and Enable System Recovery are selected, and then click Next.

5. Select Remote Shared Folder, and then click Next.

6. Type \\VAN-DC1\Backup as the remote shared folder, select Inherit, and then click Next.

7. Select VSS Full Backup, and then click Next.

8. After the backup completes, close Windows Server Backup.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 3: Performing an Authoritative Restore on a Failover Cluster In this exercise, you will delete a cluster, and then perform an authoritative restore.


• Delete a cluster.

• Perform an authoritative restore

Task 1: Delete a cluster 1. In the Failover Cluster Management snap-in, browse to VAN-CLUSTER01

\Services and Applications.

2. Delete VAN-Print.

Task 2: Perform an authoritative restore 1. Start an administrator-elevated command prompt.

2. Run wbadmin get versions to list the backups that have been completed.

3. Locate the backup that was just performed, by matching the Backup Time with the time that the backup was performed. (Note the version identifier for the backup.)

4. Run wbadmin start recovery -version:07/27/2007-19:21 -itemtype:app -items:cluster (where the version information matches the version noted in the previous step,) to start the restore.

5. After the restore is complete, run sc stop clussvc to stop the cluster service.

6. Run sc start clussvc to restart the cluster service.

7. In the Failover Cluster Management snap-in, browse to VAN-CLUSTER01 \Nodes.

8. Right-click VAN-SRV6A, and then click Resume.

9. In the Connections pane, under Nodes, click VAN-SRV6B.

10. In the Actions pane, click More Actions, and then click Start Cluster Service.

11. In the Connections pane, under Services and Applications, right-click VAN-Print, and then click Bring this service or application online.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED



window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Implementing Multi-Site Clusters 8-1

Module 8 Implementing Multi-Site Clusters

Contents: Lesson 1: Overview of Multi-Site Clusters 8-3

Lesson 2: Implementing Multi-Site Clusters Using Windows Server 2008 8-14

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

Single-site failover clusters allow applications to stay available when a server fails, or even if the cabinet that contains each node fails.

This module describes how multi-site clusters work with Windows Server®°2008, and provides general guidelines for planning a multi-site cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Multi-Site Clusters

This lesson provides information about multi-site clusters, and how they are used. Additionally, this lesson describes the two replication types that are available for creating a multi-site cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Multi-Site Clusters?

Key Points A multi-site cluster provides highly available services in more than one location. Multi-site clusters can solve a number of specific problems; however, they also present specific challenges.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: When to Use a Multi-Site Cluster


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Makes Multi-Site Clusters Unique?

Key Points In a multi-site cluster, each site usually has a separate storage system with replication between the sites. Multi-site cluster storage replication allows each site to be independent, and provides fast access to the local disk. With separate storage systems, you cannot share a single disk between sites.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Why Use a Multi-Site Cluster?

Key Points A multi-site cluster has three main advantages in a failover site compared to a remote server:

• When a site fails, a multi-site cluster automatically fails over the clustered service or application to another site.

• Because the cluster configuration is automatically replicated to each cluster node in a multi-site cluster, there is less administrative overhead than a cold standby server, which requires that you manually replicate changes.

• The automated processes in a multi-site cluster reduce the possibility of human error, which is present in manual processes.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Evaluating the Business Need for a Multi-Site Cluster

Key Points Because of increased cost and complexity of a multi-site failover cluster, it might not be an ideal solution for every application or business. When considering whether to deploy a multi-site cluster, you should evaluate the importance of the applications to the business, the type of applications, and any alternative solutions. Some applications can provide multi-site redundancy easily with log shipping or other processes, and can still achieve adequate availability with only a modest increase in cost and complexity.

The complexity of a multi-site cluster requires a higher level of architectural and hardware planning. It also requires that you develop business processes to routinely test the cluster functionality.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Storage Challenges for Multi-Site Clusters

Key Points In a multi-site cluster, Windows Server 2008 failover clustering does not include built-in functionality to replicate data between sites. There are three options for replicating data: block level hardware-based replication, software-based file replication installed on the host, or application-based replication.

Multi-site data replication can be either synchronous or asynchronous. Synchronous replication does not acknowledge data changes made in Site A until the data is successfully written to Site B. With asynchronous replication, data changes made in Site A are eventually written to Site B.

When you deploy a multi-site cluster and run the Validate a Configuration Wizard, the disk tests will not find any shared storage, and will therefore not run. If you follow the hardware manufacturer’s recommendations for Windows Server 2008 failover clustering hardware, Microsoft will support the solution.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Synchronous Replication?

Key Points When you use synchronous replication, the host receives a “write complete” response from the primary storage after the data is written successfully on both storage systems. If the data is not written successfully to both storage systems, the application must attempt to write to the disk again. With synchronous replication, both storage systems are identical.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Asynchronous Replication?

Key Points When you use asynchronous replication, the node receives a “write complete” response from the storage after the data is written successfully on the primary storage. The data is written to the secondary storage on a different schedule, depending on the hardware or software vendor’s implementation.

Asynchronous replication can be storage-based, host-based, or even application-based; however, not all forms of asynchronous replication are sufficient for a multi-site cluster. For example, Distributed File System Replications (DFS-R) provides file-level asynchronous replication; however, it does not support multi-site failover clustering replication. This is because DFS-R replicates smaller documents that are not held open continuously, and therefore, it was not designed for high-speed, open-file replication.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Comparing Synchronous and Asynchronous Replication

Key Points You use synchronous and asynchronous replication for different reasons. You use synchronous replication when data loss cannot be tolerated. Synchronous replication solutions cannot tolerate high-disk write latency, because the application waits for both storage solutions to acknowledge the data writes. The requirement for low latency disk writes also limits the distance between the storage systems because increased distance can cause higher latency. If the disk latency is high, the performance and even the stability of the application can be affected.

Asynchronous replication overcomes latency and distance limitations by acknowledging local disk writes only, and by reproducing the disk write on the remote storage system in a separate transaction. Because asynchronous replication writes to the remote storage system after it writes to the local storage system, the possibility of data loss during a failure is increased.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Network Challenges for Multi-Site Clusters

Key Points Windows Server 2008 allows cluster nodes to exist on different IP subnets, which enables a clustered application or service to change its IP address based on the IP subnet. DNS updates the clustered application’s DNS record so that clients can locate the IP address change. Because clients rely on DNS to find a service or application after a failover, you might need to adjust the DNS records’ time to live, and the speed at which DNS data is replicated. Additionally, when cluster nodes are in multiple sites, network latency might require that you modify the inter-node communication (heartbeat) delay and time-out thresholds.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Implementing Multi-Site Clusters Using Windows Server 2008

This lesson describes how to choose the correct quorum mode, and how to deploy a multi-site cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Key Windows Server 2008 Features That Enable Multi-Site Clusters

Key Points In addition to multi-subnet clustering and using inter-node communications to accommodate network latencies, Windows Server 2008 provides new cluster quorum modes (Node Majority and Node and File Share Majority) that help enable multi-site clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Quorum Issues in Multi-Site Clusters


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Choosing the Appropriate Quorum Mode in Multi-Site Clusters

Key Points Both the Node and Disk Majority, and No Majority: Disk Only quorum modes require a shared witness disk to provide a vote for determining quorum. You should only use these two quorum modes if the hardware vendor specifically recommends and supports them.

To use the Node and Disk Majority and No Majority: Disk Only modes in a multi-site cluster, the shared disk requires that:

• You preserve the semantics of the Small Computer System Interface (SCSI) commands across the sites, even if a complete communication failure occurs between sites.

• You replicate the witness disk in real-time synchronous mode across all sites.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Because multi-site clusters can have WAN failures in addition to node and local network failures, Node Majority and Node and File Share Majority are better solutions for multi-site clusters. If there is a wide area network (WAN) failure that causes the primary and secondary sites to lose communication, a majority must still be available to continue operations.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Node Majority in a Multi-Site Cluster

Key Points If you are using Node Majority and the sites lose communication, you need a mechanism to determine which nodes stay up, and which nodes drop out of cluster membership. The second site requires another vote to obtain quorum after a failure. To obtain another vote for quorum, you must join another node to the cluster, or create a file share witness.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Using Node and File Share Majority in a Multi-Site Cluster

Key Points The Node and File Share Majority mode can help maintain quorum without adding another node to the cluster. To provide for a single site failure and allow automatic failover, the file share witness might need to exist at a third site. In a multi-site cluster, a single server can host the file share witness; however, you must create a separate file share for each cluster.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Deploying Exchange Server 2007 Cluster Continuous Replication in a Multi-Site Cluster

Key Points In a Windows Server 2008 multi-site cluster, you can install Exchange Server 2007 Service Pack 1 in a Cluster Continuous Replication configuration. To keep the data on both cluster nodes synchronized without needing expensive storage replication, Exchange Server provides log shipping. You use three sites to ensure that a single site failure does not remove the ability to maintain quorum. If you cannot use three sites, you can manually move the file share witness to the secondary site, and maintain quorum during a failure.

For more information, see "High Availability"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Deploying a Multi-Site Cluster

Key Points Multi-site clusters are not appropriate for every application or every business. When you design a multi-site solution with a hardware vendor, clearly identify the business requirements and expectations. Additionally, carefully consider the quorum mode you will use, and the location of the available cluster votes.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED

Implementing Network Load Balancing Clusters 9-1

Module 9 Implementing Network Load Balancing Clusters

Contents: Lesson 1: Overview of Network Load Balancing Clusters 9-3

Lesson 2: Configuring a Network Load Balancing Cluster 9-15

Lesson 3: Maintaining a Network Load Balancing Cluster 9-30

Lab: Implementing an NLB Cluster 9-37

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Module Overview

The Network Load Balancing (NLB) feature in Windows Server®°2008 increases the availability and scalability of Internet server applications. By implementing NLB, you can combine the resources of two or more computers into a single virtual cluster that provides performance and automatic load balancing between the cluster hosts.

This module provides an overview of Network Load Balancing, and explains how to configure and maintain NLB.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 1 Overview of Network Load Balancing Clusters

NLB is a Windows Server 2008 solution that provides availability and scalability.

This lesson provides information about the NLB features, and explains how NLB operates.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Network Load Balancing (NLB)?

Key Points Applications that can benefit from being deployed on a network load balancing cluster include Web server-based applications, FTP, firewall, proxy, and virtual private network (VPN) services.

Because any server in a Network Load Balancing cluster can respond to a client request, both the application files and data on all servers must be exactly the same. Usually this means that you use a separate, back-end server to store data, or provide for a way to synchronize the data on the Web servers. However, this requirement limits which applications are suitable for load balancing.

Note: To provide a highly available system, you should consider using a failover cluster for the back-end server hosting the data.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


When a server in a Network Load Balancing cluster fails, the failure is automatically detected, and the remaining servers take over the load. In most cases, users are unaware that a server has failed.

For more information, see "Network Load Balancing Concepts"

For more information, see "Network Load Balancing Help: Overview of Network Load Balancing"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


How NLB Works

Key Points An NLB cluster uses a virtual IP address and a virtual media access control (MAC) address for network communication. All NLB nodes use the same virtual IP address and virtual MAC address. When an incoming packet is addressed to the virtual IP address and virtual MAC address, all NLB nodes receive it, but only the appropriate node responds.

NLB nodes use a common algorithm to determine which node will respond. When a client request arrives, all hosts simultaneously perform a mapping to determine which host should handle the packet. The mapping uses a randomization function that calculates a host priority based on the IP address, port, and other information. The corresponding host accepts the client request, and the other cluster hosts discard it.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


The mapping remains unchanged unless the membership of the cluster hosts changes, thus ensuring that a given client’s IP address and port will always map to the same cluster host. The algorithm load balances incoming client requests so that a selected percentage of new requests are forwarded to each cluster host. If all hosts are configured the same, the same percentage of client requests will be sent to each host. The NLB nodes do not negotiate to determine which node will respond. By using the same algorithm and monitoring, all nodes synchronize their calculations about which node should respond to each request.

When a host fails, the remaining hosts in the cluster converge, and then:

• Establish which hosts are still active members of the cluster.

• Elect the host with the highest priority as the new default host.

• Ensure that all new client requests are handled by the remaining hosts.

If the failed host begins to send heartbeat messages again, it rejoins the cluster during the convergence. After all cluster hosts agree on the current cluster membership, the client load is redistributed to the remaining hosts, and the convergence completes.

For more information, see "High Availability and Scalability Collection"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Windows Server 2008 NLB Features

Key Points All NLB components support IPv6 addresses. You can configure the addresses as the primary cluster IP address, the dedicated IP addresses, and the virtual IP addresses. In addition, you can load balance IPv6 as native IPv6 and in the IPv6 over IPv4 modes.

Network Driver Interface Specification (NDIS) 6.0 is compatible with earlier NDIS versions.

By using a callback interface, NLB can detect and notify applications during an attack, or when a node is under excessive load. Microsoft® Internet Security and Acceleration (ISA) Server uses this functionality if the cluster node is overloaded or being attacked.

NLB supports rolling upgrades from Windows Server®°2003 to Windows Server 2008.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Guide to Performing a Rolling Upgrade from a Network Load Balancing Cluster Running Windows Server 2003 to a Cluster Running Windows Server 2008"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


NLB Application Support

Key Points Applications that work well with NLB have one or more of the following characteristics:

• The applications do not require the same cluster host to service all client connections. For example, most Web sites are ideal applications for an NLB cluster because any cluster host can service the client requests. On the other hand, applications based on Secure Sockets Layer (SSL) expect that the cluster host that originated the response will maintain the SSL session.

• The applications do not require application state to be saved on a specific server in the cluster, such as session state saved by Active Server Pages (ASP). Web applications that work well on a cluster use a common database or service to maintain session state. On the initial request, a cookie is sent to the client. On subsequent requests, the client sends the cookie to the cluster. Any cluster host in the cluster can restore the application session state by using the cookie sent by the client to retrieve the session state from the session database or service.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Applications and services that are stateless are the most likely to run without modifications on NLB. Applications and services are stateless when each client request is considered to be a new request that is unrelated to any previous request.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Data Synchronization Between NLB Nodes

Key Points When you synchronize data between NLB nodes, there is always a small amount of lag-time in replication. This lag-time may cause brief inconsistent results between servers. Data synchronization is a common approach when multiple servers host a Web site, and minor discrepancies during Web site updates are acceptable.

When you store data in a common location, the responses from the NLB nodes are always consistent, because all nodes have the same view of the data. When you run Web-based applications, a common storage location is required to ensure that changes result in a consistent data structure. To do this, many Web-based applications store data in a Microsoft SQL Server® database. You must configure all hosts in the NLB cluster to communicate with a single database server, or with a clustered SQL Server.

Internet Information Services (IIS) 7.0 supports shared configuration, which enables you to store the data used by multiple Web servers on a separate file server. You can use shared configuration in combination with NLB.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Comparing NLB with Failover Clusters


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Discussion: Comparing NLB with Other Load Balancing Solutions


MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 2 Configuring a Network Load Balancing Cluster

In order to implement NLB in Windows Server 2008, you must install the NLB feature, and then create and configure NLB clusters using the Network Load Balancing Manager. This lesson describes how to set up and configure NLB clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


NLB Deployment Requirements

Key Points We recommend that all hosts in an NLB cluster have two network adapters. This simplifies network communication by isolating NLB cluster traffic from host traffic. Although NLB in Windows Server 2008 supports a single network adapter for both unicast and multicast cluster operation modes, multiple network adapters provide enhanced redundancy and performance.

You must configure the your network adapter for NLB with TCP/IP, and all NLB nodes must be on the same subnet. This is required because all nodes in the cluster share a single virtual IP address. If the cluster hosts are not connected to the same physical subnet, ensure that they are connected to a virtual subnet.

If the load balancing cluster is operating close to capacity, the loss of a server may increase the load on remaining cluster nodes, and subsequently degrade client performance. To avoid performance degradation due to a failed server, ensure that each host in the load balancing cluster has sufficient capacity to handle the added load.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Creating NLB Clusters

Questions:

What are the prerequisites for creating the NLB cluster?

How is creating an NLB cluster different than creating a failover cluster?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Configuring NLB Clusters and Hosts

Key Points The cluster IP address is the virtual IP address that is assigned to the cluster. Client requests are sent to the cluster IP address.

When you enter the fully qualified domain name (FQDN) in the cluster parameters, the name is not automatically registered in a Domain Name System (DNS), or used by other Windows components. Not assigning a full Internet name setting does not affect the operation of any Windows component, including NLB.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


As you plan the NLB cluster and host settings, the configuration choices you make determine the following:

Choice Configuration settings

Will each server host the same number of client connections?

Determines the port rule load weight setting

Will the client connect to the same server for all connections?

Determines the affinity setting

Will the cluster operate the same way for all applications?

Determines the port rule settings

Do network switches support unicast or multicast operations?

Determines the cluster operation modes

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Comparing Unicast and Multicast NLB Cluster Operation Modes

Key Points We recommend that you select the unicast method for distributing client requests, unless only one network adapter is installed in each cluster host.

Considerations for using unicast mode include:

• In order to provide peer-to-peer communication between cluster hosts, a second network adapter is required.

• If you connect the cluster to a switch, incoming packets are sent to all the ports on the switch, which may cause switch flooding. Outgoing packets use a modified version of the cluster MAC address that is unique to each node. This prevents switches from associating the cluster MAC address to a specific switch port.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for using the multicast mode include:

• Upstream routers might require a static Address Resolution Protocol (ARP) entry. This is because routers might not accept an ARP response that resolves unicast IP addresses to multicast MAC addresses.

• The Internet Group Management Protocol (IGMP) multicast check box enables IGMP support for limiting switch flooding, by limiting traffic to Network Load Balancing ports only. Enabling IGMP support ensures that traffic intended for an NLB cluster passes through only those ports serving the cluster hosts, and not all switch ports. Without IGMP, switches might require additional configuration to tell the switch which ports to use for the multicast traffic.

• Upstream routers might not support mapping a unicast IP address (the cluster IP address) with a multicast MAC address. In these situations, you must upgrade or replace the router, or use the unicast mode.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Options for Configuring Host Parameters

Key Points For each NLB node, you can configure host parameters that are specific to that node. The host parameters are:

• Priority: Each NLB node must have a unique priority value. If the port rule does not address traffic addressed to the cluster, the NLB node with the lowest-priority value will respond. If you specify the same cluster host priority for two cluster hosts, the last cluster host that starts will not join the cluster, and an error message that describes the problem is sent to the Windows system event log. The existing cluster hosts continue to operate as before.

• Dedicated IP address: This specifies the IP address that is used on the host for cluster management or general computer management. NLB ensures that port rules do not affect traffic to the dedicated address. You must list the dedicated IP address first in the list of IP addresses, in the TCP/IP properties.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Subnet mask.

• Initial host state: Usually, you want the host to join automatically so that can begin working after a reboot, with no manual intervention. However, you may not want it to join automatically if you are performing maintenance on the host and will be rebooting several times as part of the maintenance process.

For more information, see "Specifying the Cluster Host Parameters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Are Port Rules?

Key Points Port rules specify how requests to a certain IP address and port range are directed to cluster hosts. This enables you to define different rules for different applications running on the NLB cluster. For example, you might want to distribute requests for a Web application on TCP port 80 evenly among servers, but direct all requests for the Web application on TCP port 8080 to a single server.

Windows NLB uses the filtering mode in a port rule to determine how to distribute requests among nodes in the NLB cluster. The options for filtering mode are:

• Multiple hosts: All NLB nodes respond based on the weight assigned to each node. The weight for each node is assigned depending on the physical capabilities of the node. If a node fails, then other nodes continue to respond. Distributing requests among multiple nodes increases the availability and scalability of an application.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Single host: Only the NLB node with the highest priority responds. If the highest-priority node fails, then the next highest-priority node begins to respond. Distributing requests to only a single host increases availability, but does not increase scalability.

• Disable this port range: All packets for this port range are blocked. Use this option when the cluster does not run any applications on a specific port range.

Important: The port rules that you apply to each cluster host must be identical, with the exception of the load weight (in the multiple hosts filter mode,) and the handling priority (in the single hosts filter mode). If there is a discrepancy between port rules on existing cluster hosts, the cluster will not converge.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


What Is Affinity?

Key Points The affinity option controls distribution of requests from a specific client among the nodes in an NLB cluster. This setting is only relevant when you select the multiple-hosts filtering mode. The options for affinity are:

• None: Any node can respond to any client request. This is suitable for stateless applications such as static Web pages. The first Web page can be retrieved from Node A, and the second Web page from Node B.

• Single: A single node responds to all requests from a single client. This is required for state-based applications such as Web applications that you must authenticate. The authentication state is only stored on the server that you originally contacted. If requests are answered by any node other than the authenticating node, the request is unsuccessful, or the client is forced to authenticate again.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


• Network: A single node responds to all requests from a class C network. This is useful when clients are accessing the NLB cluster from behind load-balanced proxy servers. Requests coming from each proxy server will have a different source IP address, but will be on a single, class C network. This option ensures that a connection can be maintained even when the source IP address may vary.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Configuring an NLB Cluster

Questions:

You want one host in the NLB cluster to handle all traffic sent to a Web application using port 8888. How would you configure the cluster parameters?

You have deployed a new server in an existing NLB cluster and want to be sure that this server will service twice as many client connections as the existing servers. How would you configure the NLB cluster?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Configuring NLB

Key Points When you use Network Load Balancing Manager to configure cluster parameters, you enter the cluster parameters once during the creation of the cluster. As you add cluster hosts to the cluster, Network Load Balancing Manager automatically configures the cluster parameters for the new cluster hosts.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lesson 3 Maintaining a Network Load Balancing Cluster

After deployment, you must manage the NLB cluster by modifying the cluster settings, and using the Network Load Balancing Manager to manage hosts in the cluster. This lesson describes how to manage and troubleshoot NLB clusters.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Managing NLB Clusters

Questions:

What happens when you delete a host from the Network Load Balancing Manager?

After you enable logging in the Network Load Balancing Manager, what types of events will show up in the NLB log file?

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Demonstration: Managing Hosts on an NLB Cluster

Question:

You must remove a host from an NLB cluster. What should you do before you remove the host so that you do not disrupt client connections?

Note: You can perform many of these tasks by using the command-line tool NLB.exe. For details, see "Controlling Hosts on Network Load Balancing Clusters" in Network Load Balancing Online Help.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Tools for Troubleshooting NLB Clusters

Key Points Troubleshooting an NLB cluster usually falls into one of two categories: client connectivity issues, and NLB load balance issues. The following tools help you to troubleshoot an NLB cluster:

• Use the standard network troubleshooting tools on the client, and use Network Monitor on the servers to troubleshoot client connectivity issues.

• Monitor the events in the System Log in Event Viewer to troubleshoot NLB cluster issues such as cluster convergence issues, or server performance issues.

• Use the Reliability and Performance Monitor to troubleshoot performance issues.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


For more information, see "Network Load Balancing Online Help: Troubleshooting Network Load Balancing Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Considerations for Troubleshooting NLB Clusters

Key Points When troubleshooting an NLB cluster, consider the following issues:

Issues Symptoms Resolution

Network connectivity issues

• Duplicate IP address error

• Host unreachable messages with Ping or Network Load Balancing Manager

• No response in a management tool

• Ensure that each host has a unique dedicated address.

• Ensure that cluster operation mode is the same on all hosts.

• Ensure that the correct network adapters are assigned to the cluster IP address.

• Verify network support for multicasting, if enabled.

• Check firewall configurations.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


(continued)

Issues Symptoms Resolution

Application connectivity issues

• Some users gain access while others do not.

• TCP connections are reset.

• Verify that applications are installed correctly.

• Verify application authorization settings.

• Verify that the cluster IP address is added to all hosts.

• Disable Hypertext Transfer Protocol (HTTP) keep-alives.

• Monitor server performance.

NLB convergence issues

• Hosts start, but never finish convergence.

• Hosts move in and out of converged state.

• Verify that the same port rules are defined on all hosts.

• Verify network connectivity on the cluster network.

• Verify that all cluster nodes are running Windows Server 2008.

Load balancing issues

• Application load is not distributed evenly.

• Only the default host is responding to client requests.

• Verify network connectivity between hosts.

• Verify that all hosts are on the same subnet.

• Check the port rule configuration.

• Verify that client connections are not coming from a single IP address, (that is, through a proxy server.)

For more information, see "Reference: Network Load Balancing Troubleshooting"

For more information, see "Network Load Balancing Help: Troubleshooting Network Load Balancing Clusters"

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Lab: Implementing an NLB Cluster

Lab Objectives • Prepare the NLB cluster nodes.

• Configure an NLB failover cluster.

Scenario You are the server administrator for Fabrikam Inc., and are implementing a new Web-based application. To increase the availability of this new application, you will use NLB in Windows Server 2008.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 1: Preparing the NLB Cluster Nodes You must install the new Web-based application on servers VAN-SRV3A and VAN-SRV3B. In this exercise, you will configure the Web server role on each server before you configure and test the Web application on each server. You will also install the Network Load Balancing role.


• Start the required virtual machines.

• Prepare network connections for NLB on VAN-SRV3A.

• Prepare network connections for NLB on VAN-SRV3B.

• Create a Web site on VAN-SRV3A.

• Create a Web site on VAN-SRV3B.

• Configure firewall rules for the Web site.

• Verify Web site functionality.

• Install the Network Load Balancing feature.

Task 1: Start the required virtual machines 1. On your host machine, click Start, point to All Programs, point to Microsoft






6. Log on to all computers as Administrator using a password of Pa$$w0rd.

Task 2: Prepare network connections for NLB on VAN-SRV3A • On VAN-SRV3A, rename the network connections:

• Local Area Connection: Client Network

• Local Area Connection2: Cluster Network

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 3: Prepare network connections for NLB on VAN-SRV3B • On VAN-SRV3B, rename the network connections:

• Local Area Connection: Client Network

• Local Area Connection2: Cluster Network

Task 4: Create a Web site on VAN-SRV3A 1. On VAN-SRV3A, create the folder C:\Webapp.

2. Copy the file C:\Inetpub\SVR3A.txt to C:\Webapp\default.htm.

3. Use the IIS Manager administrative tool to create a new Web site with the following settings:

• Web site name: Webapp

• Physical path: C:\Webapp

• Port: 8080

Task 5: Create a Web site on VAN-SRV3B 1. On VAN-SRV3B, create the folder C:\Webapp.

2. Copy the file C:\Inetpub\SVR3B.txt to C:\Webapp\default.htm.

3. Use the IIS Manager administrative tool to create a new Web site with the following settings:

• Web site name: Webapp

• Physical path: C:\Webapp

• Port: 8080

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 6: Configure firewall rules for the Web site 1. On VAN-SRV3A, use the Windows Firewall with Advanced Security

administrative tool to create a new inbound rule with the following attributes:

• Rule type: Port

• Port type: TCP

• Specific local port: 8080

• Allow the connection

• Profiles: all

• Name: Web Application

2. On VAN-SRV3B, use the Windows Firewall with Advanced Security administrative tool to create a new inbound rule with the following attributes:

• Rule type: Port

• Port type: TCP

• Specific local port: 8080

• Allow the connection

• Profiles: all

• Name: Web Application

Task 7: Verify Web site functionality • On VAN-DC1, use Internet Explorer to view the following Web sites:

• http://VAN-SRV3A:8080

• http://VAN-SRV3B:8080

Task 8: Install the Network Load Balancing feature 1. On VAN-SRV3A, use Server Manager to install the Network Load Balancing

feature.

2. On VAN-SRV3B, use Server Manager to install the Network Load Balancing feature.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Exercise 2: Configuring an NLB Failover Cluster To increase the availability of the new Web-based application, you decide to create an NLB cluster with failover. In this exercise, you will create an NLB failover cluster and a port rule for the Web-based application.


• Create an NLB cluster.

• Add VAN-SRV3B to the NLB cluster.

• Configure a Domain Name System (DNS) record for the NLB cluster.

• Configure a port rule for failover.

• Verify cluster failover.

• Configure a port rule for load balancing.

• Verify cluster load balancing.

Task 1: Create an NLB cluster 1. On VAN-SRV3A, use the Network Load Balancing Manager administrative tool

to create a new cluster.

• Connect to VAN-SRV3A.

• Use the Client Network interface.

• Accept the default host parameters.

• Cluster IP address: 10.10.0.100

• Subnet mask: 255.255.0.0

• Full Internet name: webapp.fabrikam.com

• Operation mode: Unicast

2. Verify that the 10.10.0.100 address has been added to the Client Network interface.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Task 2: Add VAN-SRV3B to the NLB cluster 1. On VAN-SRV3B, use the Network Load Balancing Manager administrative tool

to connect to the existing cluster webapp.fabrikam.com on VAN-SRV3A.

2. Add VAN-SRV3B as a node to the cluster.

3. Verify that the 10.10.0.100 address has been added to the Client Network interface.

Task 3: Configure a Domain Name System (DNS) record for the NLB cluster • On VAN-DC1, in DNS, create a new host record for WebApp.fabrikam.com, in

the Fabrikam.com zone. Configure an IP address of 10.10.0.100.

Task 4: Configure a port rule for failover 1. On VAN-SRV3A, use the Network Load Balancing Manager administrative tool

to open the cluster properties.

2. On the Port Rules tab, edit the existing rule with the following settings:

• Port range: from 8080 to 8080

• Protocols: TCP

• Filtering mode: Single host

Task 5: Verify cluster failover 1. On VAN-DC1, use Internet Explorer to view the Web site on the NLB cluster,

http://webapp.fabrikam.com:8080.

2. Reload the page several times to confirm the server from which the page is being loaded.

3. On VAN-SRV3A, in the Load Balancing Manager administrative tool, stop the server from which the page was being loaded by using the Control Host submenu.

4. On VAN-DC1, use Internet Explorer to view the Web site on the NLB cluster, http:// webapp.fabrikam.com:8080.

MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


5. Verify that the Web page loaded from the other server.

6. On VAN-SRV3A, in the Load Balancing Manager administrative tool, start the stopped node by using the Control Host submenu.

Task 6: Configure a port rule for load balancing 1. On VAN-SRV3A, use the Network Load Balancing Manager administrative

tool to open the properties of the cluster.

2. On the Port Rules tab, edit the existing rule with the following settings:

• Filtering mode: Multiple host

• Affinity: None

Task 7: Verify cluster load balancing 1. On VAN-DC1, use Internet Explorer to view the Web site on the NLB cluster,

http://webapp.fabrikam.com:8080.

2. On VAN-SRV3A, use Internet Explorer to view Web site on the NLB cluster, http://webapp.fabrikam.com:8080.

3. On VAN-Web, use Internet Explorer to view Web site on the NLB cluster, http://webapp.fabrikam.com.:8080.

4. Verify that the Web page loads from both Web servers.

Task 8: Close all virtual machines, and discard undo disks 1. For each running virtual machine, close the Virtual Machine Remote Control

window.



MC

T USE O

NLY. STU

DEN

T USE PR

OH

IBITED


Course Evaluation

Your evaluation of this course will help Microsoft understand the quality of your learning experience.

Please work with your training provider to access the course evaluation form.

Microsoft will keep your answers to this survey private and confidential and will use your responses to improve your future learning experience. Your open and honest feedback is valuable and appreciated.

Documents

MOC Course 6423A Implementing & Managing Windows Server 2008 Clustering