Interconnecting Data Centers Using VPLS (Ensure Business Continuance on Virtualized Networks by Implementing Layer 2 Connectivity Across Layer 3)

Product Description

As data centers grow in size and complexity, enterprises are adopting server virtualization technologies such as VMware, VMotion, NIC teaming, and server clustering to achieve increased efficiency of resources and to ensure business resilience. However, these technologies often involve significant expense and challenges to deal with complex multisite interconnections and to maintain the high availability of network resources and applications.

Interconnecting Data Centers Using VPLS presents Virtual Private LAN Service (VPLS) based solutions that provide high-speed, low-latency network and Spanning Tree Protocol (STP) isolation between data centers resulting in significant cost savings and a highly resilient virtualized network. The design guidance, configuration examples, and best practices presented in this book have been validated under the Cisco Validated Design (CVD) System Assurance program to facilitate faster, more reliable and more predictable deployments. The presented solutions include detailed information about issues that relate to large Layer 2 bridging domains and offer guidance for extending VLANs over Layer 3 networks using VPLS technology.

Implementing this breakthrough Data Center Interconnect (DCI) strategy will evolve your network to support current server virtualization techniques and to provide a solid foundation for emerging approaches. The book takes you from the legacy deployment models for DCI, problems associated with extending Layer 2 networks, through VPN technologies, to various MST-, EEM-, and GRE-based deployment models and beyond. Although this book is intended to be read cover-to-cover, it is designed to be flexible and allow you to easily move between chapters to develop the solution most compatible with your requirements.

• Describes a variety of deployment models to effectively transport Layer 2 information, allowing your virtualization solution to operate effectively
• Explains benefits and trade-offs of various solutions for you to choose the solution most compatible with your network requirements to ensure business resilience
• Provides detailed design guidance and configuration examples that follow Cisco best practice recommendations tested within the CVD

This book is part of the Networking Technology Series from Cisco Press®, which offers networking professionals valuable information for constructing efficient networks, understanding new technologies, and building successful careers.

Sample Pages

Download the sample pages (includes Chapter 1 and Index)

Table of Contents

Introduction xv

Overview of High-Availability Clusters 2

Public Network Attachment 3

Private Network Attachment 3

Data Center Interconnect: Legacy Deployment Models 4

Problems Associated with Extended Layer 2 Networks 5

Summary 7

VPN Technology Considerations 9

Layer 3 Virtual Private Networks 10

Layer 2 Virtual Private Networks 10

VPLS Overview 11

Understanding Pseudowires 14

VPLS to Scale STP Domain for Layer 2 Interconnection 15

H-VPLS Considerations 17

EEM 18

MPLS 19

Label Switching Functions 19

MPLS LDP 20

Summary 22

MTU Evaluation for Intersite Transport 23

Core Routing 25

Mixed MPLS/IP Core 26

Pure MPLS Core 28

Pure IP Core 30

Convergence Optimization 32

Key Convergence Elements 33

Failure Detection and Tuning 33

Alternate Route Computation 40

Summary 42

Understanding MPLS-TE 43

Fast Reroute 44

Load Repartition over the Core 45

Load Repartition over a Parallel-Links Bundle 45

Implementing MPLS-TE for Traffic Repartition over Parallel Links 46

Enable TE 47

Create MPLS-TE Tunnels and Map Each VFI to a Tunnel LSP 48

Explicit-Path Option 48

Affinity Option 52

Summary 53

Ensuring a Loop-Free Global Topology: Two Primary Solution Models 55

N-PE Using MST for Access to VPLS 56

N-PE Using ICCP Emulation for Access to VPLS 56

Data Center Interconnect Design Alternatives: Summary and Comparison 57

Case Study 1: Large Government Organization 61

Challenges 61

Solution 62

Case Study 2: Large Outsourcer for Server Migration and Clustering 65

Challenges 65

Solution 65

Summary 68

Chapter 7 Data Center Multilayer Infrastructure Design 69

Network Staging for Design Validation 71

Hardware and Software 72

Convergence Tests 73

Summary 76

MST in N-PE: MST Option 1a 77

Implementing MST in N-PE: MST Option 1a Design 80

Convergence Tests 100

Cluster Server Tests 103

VPLS with N-PE Redundancy Using RPVST with Isolated

MST in N-PE: MST Option 1b 106

EEM Scripting to Complement Isolated MST Solution 109

Implementing RPVST in a Data Center with Isolated MST

in N-PE (MST Option 1b) Design 110

Convergence Tests 130

Cluster Server Tests 134

Summary 138

N-PE Redundancy Using the Semaphore Protocol: Overview 139

Semaphore Definition 141

Semaphore Theory Application 142

N-PE Redundancy Using Semaphore Protocol: Details 142

VPLS PWs in Normal Mode 142

Primary N-PE Failure 145

Primary N-PE Recovers After the Failure 145

Implementing a Semaphore 146

EEM / Semaphore Scripts 147

Naming Conventions 148

VPLS with N-PE Redundancy Using EEM Semaphore:

EEM Option 2 150

Control Plane 151

Data Plane 151

Theory of Operation 151

N-PE Routers: Hardware and Software 154

Implementing VPLS with N-PE Redundancy Using

EEM Semaphore Design 154

Convergence Tests 168

Cluster Server Tests 172

H-VPLS with N-PE Redundancy Using EEM Semaphore:

EEM Option 3 176

Control Plane 179

Data Plane 179

Theory of Operation 179

N-PE Routers: Hardware and Software 182

Implementing H-VPLS with N-PE Redundancy Using EEM Semaphore

Design 182

Convergence Tests 195

Server Cluster Tests 199

Multidomain H-VPLS with N-PE Redundancy: EEM Option 4a 201

Control Plane 203

Data Plane 203

Theory of Operation 204

N-PE Routers: Hardware and Software 207

Implementing Multidomain H-VPLS with N-PE Redundancy

Using EEM Semaphore Design 207

Convergence Tests 217

Server Cluster Tests 221

Multidomain H-VPLS with Dedicated U-PE: EEM Option 4b 227

Multidomain H-VPLS with Multichassis EtherChannel: EEM Option 5a 227

Solution Positioning 230

Multidomain H-VPLS with MEC and VLAN Load Balancing:

EEM Option 5b 230

Control Plane 233

Data Plane 233

Theory of Operation 233

N-PE Routers: Hardware and Software 236

Implementing EEM Option 5b 237

Convergence Tests 252

Server Tests 259

Multidomain H-VPLS with MEC and VLAN Load Balancing:

PWs on Active and Standby VPLS Nodes in Up/Up State:

EEM Option 5c 262

N-PE Routers: Hardware and Software 264

Configuration Summary 265

Convergence Tests 270

Summary 275

Key Configuration Steps for VPLSoGRE-Based Solutions 279

VPLSoGRE with N-PE Redundancy Using EEM Semaphore 282

Convergence Tests 284

Cluster Server Tests 286

VPLSoGRE: Multidomain with H-VPLS Solution 291

Convergence and Cluster Server Tests 296

Cluster Server Tests 298

Summary 302

Multicast Deployment in a Layer 2 Environment 303

Multicast at Layer 2 304

Tuning the IGMP Query Interval 304

Spanning Tree, HSRP, and Service Module Design 306

Routing Design 306

QinQ MAC Overlapping 307

Storm Control 310

L2 Control-Plane Packet Storm Toward N-PE 311

L2 Broadcast and Multicast Packet Storm 312

L2 Known Unicast Packet Storm 313

L2 Unknown Unicast Packet Storm 314

QoS Considerations 315

Stateful Switchover Considerations 318

IGP (OSPF) Cost 318

Router ID Selection 319

Summary 319

Introducing ICCP 322

Interaction with AC Redundancy Mechanisms 324

Interaction with PW Redundancy Mechanisms 325

Configuring VPLS PE Redundancy Using ICCP 326

Summary 327

A Larger Problem to Solve 329

Networking Technology: Research Directions 330

Improving Legacy L2 Bridging 330

New Concepts in L2 Bridging 331

L2 Service over L3 Transport: MPLS or IP? Battle or Coexistence? 332

Summary 333