Microsoft Hyper-V Cluster Design电子书

Microsoft Hyper-V Cluster Design

Table of Contents

Microsoft Hyper-V Cluster Design

Credits

About the Author

About the Reviewers

www.PacktPub.com

Support files, eBooks, discount offers and more

Why Subscribe?

Free Access for Packt account holders

Instant Updates on New Packt Books

Preface

What this book covers

What you need for this book

Who this book is for

Conventions

Reader feedback

Customer support

Downloading the example code

Errata

Piracy

Questions

1. Hyper-V Cluster Orientation

Terminology

Clustering in a Microsoft environment

Create a project document

Purposes for a Hyper-V Server cluster

High availability

High Availability Printing

Balancing resources

Geographic dispersion

Natural replacement for aging infrastructure

Test, development, and training systems

Cloud hosting

Resource metering

VDI and RemoteFX

Be open to other purposes

Goals for a Hyper-V Server cluster

Identify the resources that cannot be virtualized

Consult with application vendors

Involve internal stakeholders

Define phases and timelines

Perform further research

Define success metrics

Measure and predict your workload

Only allow changes during the planning phase

Looking forward to the Design phase

Host computers

Storage

Cluster Shared Volumes

SMB shares

Mixing SMB 3.0 and CSV

Networking

Management

Cluster and Cluster Shared Volumes

Live Migration

Subnetting

Virtual machine traffic

Storage traffic

Physical adapter considerations

Adapter teaming

Active Directory

Virtualized domain controllers

Supporting software

Management tools

Backup

Training

A sample Hyper-V Cluster planning document

Sample project title – Techstra Hyper-V Cluster Project

Sample project – purposes

Sample project – goals

Sample project – success metrics (subsection of goals)

Review the sample project

Summary

2. Cluster Design and Planning

Starting the design phase

Planning for existing systems

Deciding how you will virtualize physical systems

Determining requirements for existing systems

Microsoft Assessment and Planning Toolkit

Performance Monitor

General approaches to reading the metrics

Memory measurements

Network measurements

Disk measurements

Processor measurements

Host computer components

Hyper-V Server requirements

CPU

Memory

Host networking

Host storage

Management operating system

Hyper-V Server

Windows Server

Deciding on a management operating system

Deciding between Hyper-V Server 2012 and 2012 R2

Networking

Advanced networking hardware

Shared storage

Storage area network devices

Network-attached storage devices

General purpose computers

Shared storage performance characteristics

Designing shared storage

Software licensing

Windows Server and guest virtualization rights

Software Assurance

Client access licenses

Other software licenses

Hyper-V and cluster-related software planning

Remote software applications

Local software applications

Blade hardware

Physical placement

Security

Domain separation

Hyper-V isolation

Network isolation

Complete the planning phase

Sample project – planning and design

Sample project – hardware

Summary

3. Constructing a Hyper-V Server Cluster

Documenting the initial setup phase

Build steps not covered in this book

Auxiliary built-in tools

Acquiring and enabling the GUI tools

Enabling the tools on Windows 8/8.1 from the GUI

Enabling the tools on Windows Server 2012/R2 in the GUI

Enabling the tools using PowerShell

Configuring nodes

Initial node configuration using GUI tools

Using the GUI to configure networking

Renaming network adapters

Creating network teams

Enable roles and features

Creating or modifying a virtual switch

Creating virtual adapters for converged fabric

Setting IP addresses for management operating system adapters

Joining the computer to the domain

Initial node configuration using PowerShell

Basic configuration

Enable roles and features

Using PowerShell to configure networking

Rename network adapters

Converged fabric

Creating network teams

Create a virtual switch

Create virtual adapters

Assigning virtual adapters to VLANs

Setting IP addresses for management operating system adapters

Join the computer to the domain

Optional node configuration steps

Prepare other nodes

Building the cluster

Cluster validation

Running cluster validation in the GUI

Using PowerShell for cluster validation

Cluster creation

Creating a cluster using the GUI

Creating a cluster using PowerShell

Handling cluster creation errors

Cluster post-creation steps

Prepare storage

Add prepared storage

Configure quorum

Configure networks

Set Live Migration network preferences

Summary

4. Storage Design

Early storage planning

Physical storage characteristics

Physical disks

Drive bus

Traditional RAID

Storage Spaces

Shared storage connectivity

Hyper-V Server storage space utilization

Management operating system

BIN files

VSV files

XML files

SLP files

VFD files

VHD and VHDX files

IDE and SCSI virtual controllers

VHD versus VHDX

VHD and VHDX types

Fixed

Dynamic

Differencing

Fixed versus dynamic disks

Pass-through disks

Other storage usage considerations

Hyper-V Server storage performance

Pass-through disks

Expansion

Fragmentation

Fragmentation and dynamic VHDX performance

Working with storage

Connecting to iSCSI storage

Connecting to Fibre Channel storage

Connecting to SMB 3.0 shares

Enabling and using Multipath IO

Managing disks

Cluster Shared Volumes

Finding and renaming Cluster Shared Volumes

CSV cache

Placing virtual machines on storage

Storage deduplication

Storage QoS (2012 R2 Only)

Enhanced features

Summary

5. Network Design

The Hyper-V virtual switch

Network virtualization

Redundancy and load balancing for the virtual switch

Assign virtual adapters to VLANs

Hyper-V Server networking in a cluster

Management

Redundancy and load balancing for management traffic

Cluster communications

Redundancy and load balancing for cluster communications

SMB multichannel configuration

Live Migration

Redundancy and load balancing for Live Migration traffic

Setting the Live Migration mode in 2012 R2

Storage connectivity

The virtual switch in a cluster

Adapter teaming

Teaming fundamentals

Teaming modes

Switch Independent teaming

Static teaming

LACP teaming

Load balancing algorithms

Address Hash techniques

Transport Ports hash

IP Addresses hash

MAC Addresses hash

Hyper-V Port balancing

Dynamic (R2 only)

The effects of teaming selections

Switch Independent with hash

Switch Dependent (static or LACP) with hash

Switch Independent with Hyper-V Ports

Switch Dependent (static or LACP) with Hyper-V Ports

Understanding the change brought by R2's Dynamic algorithm

Effect of teaming on other technologies

Practical teaming guidance

Converged fabric

Practical converged fabric guidance

Planning the physical layout

Firewall settings

Remote desktop

PowerShell

Firewall rules

Summary

6. Network Traffic Shaping and Performance Enhancements

Windows Server Quality of Service

Policy-based QoS

Hyper-V QoS

Data Center Bridging

802.1p tagging

Assigning applications and traffic types to QoS classes

Setting bandwidth on DCB QoS classes

The default class

Changing advanced settings on network adapters

Advanced adapter settings in the GUI

Advanced adapter settings in the registry

Advanced adapter settings in PowerShell

Jumbo frames

VMQ

VMQ and adapter teaming

VMQ interrupt coalescing

RSS

vRSS (R2 only)

RSS tuning

RDMA

SR-IOV

SR-IOV tuning

Other hardware-assisted offloading technologies

Virtual adapter networking control

Practical advice for network performance design

Further possibilities

Summary

7. Memory Planning and Management

Understanding physical memory characteristics

Memory types

U-DIMM

R-DIMM

FB-DIMM

LR-DIMM

Memory speed

Practical guidance on memory speeds

Memory ranks

Mirroring, sparing, ECC, and other options

Practical memory protection

Memory channels

Practical multi-channel memory implementation

NUMA

Practical NUMA configuration

Physical memory installation

How Hyper-V Server uses memory

Host memory

Hypervisor memory usage

Hyper-V Server and NUMA

Virtual machine memory

Practical virtual machine memory sizing

Virtual machines and NUMA

NUMA in a cluster

Practical virtual machine NUMA configuration

Dynamic Memory

Dynamic Memory requirements and restrictions

Startup RAM, Minimum RAM, and Smart Paging

Practical virtual machine memory implementation

Cluster memory shortages

Summary

8. Performance Testing and Load Balancing

Initial and on-going performance measurement

General performance measurement

Server Performance Advisor

Performance Monitor

Real-time monitoring with Performance Monitor

Trend tracking with Performance Monitor

Selecting counters practically

Alternative ways to read performance logs

Subsystem testing

Disk I/O testing

Practical IOMeter usage for disk analysis

Network testing

IOMeter for network testing

Practical IOMeter usage for network analysis

NTttcp for network testing

Memory testing

Baseline and comparative performance measures

Cluster load balancing

Preferred owners

Setting preferred owners using Failover Cluster Manager

Setting preferred owners using PowerShell

Possible owners

Setting possible owners using Failover Cluster Manager

Setting possible owners using PowerShell

Anti-affinity

Summary

9. Special Cases

Non-highly-available virtual machines in a cluster

Local virtual machines

Switching to or from high availability mode

Converting a local virtual machine to high availability

Restricted highly available virtual machines

A cluster with only one virtual machine

Single-VM cluster in a small environment

A cluster with a single host

Virtualized domain controllers in a Hyper-V Cluster

Discomfort with virtualization of a vital infrastructure role

Concern that Hyper-V Server will not start

Concern that domain controllers will be unavailable

Concern over clock drift in a virtual environment

Concern over effects of snapshots on domain controllers

Concerns over Saved States of domain controllers

Security concerns for virtualized domain controllers

Implementing virtualized domain controllers in a cluster

Windows domain time synchronization

Storing a single virtual machine's files in different locations

Geographically distributed clusters

Cluster networking with multiple subnets

Cluster name object in a multiple subnet cluster

Configuring subnet traffic handling

Virtual machine networking in a multiple subnet cluster

Using non-virtualized hardware in a cluster

Pass-through disks in a cluster

Summary

10. Maintaining and Monitoring a Hyper-V Server Cluster

Cluster validation

When to perform validation

Running the validation wizard in Failover Cluster Manager

Validating a cluster in PowerShell

Reading the validation report

Other cluster reports

Best Practices Analyzer for Hyper-V

Hyper-V Best Practices Analyzer in Server Manager

Hyper-V Best Practices Analyzer in PowerShell

Updating Hyper-V Server hosts

Cluster-Aware Updating

Configuring Cluster-Aware Updating

Enabling CAU using the GUI

Enabling CAU using PowerShell

Windows Update and Windows Server Update Services

Hotfixes

Monitoring Hyper-V Server

Event logs

Services

Metering

Summary

11. High Availability

What does high availability mean?

Service level agreements

Fault tolerance

Creating a highly available virtual machine

Using Failover Cluster Manager to create a highly available virtual machine

Using Failover Cluster Manager to make an existing virtual machine highly available

Using PowerShell to create and convert a highly available virtual machine

Removing high availability from a virtual machine

High availability beyond Hyper-V

Cluster within a cluster

Shared VHDX in R2

Network adapter configurations for guest clusters

Migrations

Practical high availability migration guidance

Cluster responses to failures

Automatic Stop Action

Automatic Start Action

Failback

Failover limits

Network protection in R2

Quorum

Configuring quorum using Failover Cluster Manager

Configuring quorum using PowerShell

R2-only quorum PowerShell operations

Practical quorum guidance

Recovering from quorum loss

Snapshots or checkpoints

Practical snapshot guidance

Summary

12. Backup and Disaster Recovery

Knowing your risks

Physical loss

Data loss

Data corruption

Risk analysis

Risk mitigation

Planning for disaster recovery

Backup

Choosing a backup solution

Architecting a backup solution

Choosing what to back up

Choosing when to back up

Deciding how long to keep backup data

Hybrid solutions

Storage for backup

Deploying a backup solution

Windows Server Backup

Hyper-V Replica

Architecting a Hyper-V Replica solution

Choosing where to send replicas

Choosing what to replicate

Deciding how many recovery points to keep

Deploying Hyper-V Replica

Practical Hyper-V Replica guidance

Verifying your disaster recovery plan

Summary

Index