High Availability Networking
with Cisco

Vincent C. Jones

Table of Contents (Detailed)

Front Matter


1. Reliability Theory.
The Need for Network Availability.
   Defining Network Availability
   Determining Availability Requirements
   Availability Measurement and Reporting
Mathematics of Reliability and Availability.
   MTBF and MTTR
   Improving Availability through Higher Reliability
   Improving Availability through Redundancy
   Need for Testing
   Functional Availability
Network Vulnerabilities.
   Classical Hardware Considerations
   Software Considerations
   Site Infrastructure
   Communications Links
Designing for Improved Network Availability.
   WAN Communications Diversity
   Internal Diversity

2. Bridging and Routing.
Basic Connectivity Considerations.
   Connectivity in the OSI Reference Model
   Network Topologies
   Multiple Connections versus Single Connections
   Detecting Failure
Bridging for Higher Availability.
   Classic Bridging
   Token Ring Source Route Bridging
   Transparent Spanning-Tree Bridging
   Switched LAN Design
Routing for Higher Availability.
   Routing Fundamentals
   Static Routing
   RIP Routing
   OSPF Routing
   EIGRP Routing
   Integrated IS-IS Routing
   BGP Routing
   Routed Network Design

3. Multihomed Hosts.
Two Network Interfaces with Independent Addresses.
   Impact of Network Connection Failure
   Application Recovery Requirements
Two Network Interfaces with One IP Address.
   Smart Network Interface Cards
   Anatomy of a Novell Server
   Configuration Example: Generic UNIX using routed
   Configuration Example: Practical Linux Server using GateD
Two Systems with One IP Address.
   Server Clusters
   Server Cluster Considerations
   Server Farms
   Intelligent Clients

4. Dial Backup for Permanent Links.
General Considerations.
   Getting Adequate Bandwidth on Demand
   Coping with Backup Bandwidth Limitations
   The Need for Constant Testing
Dial Backup Approaches.
   Dialing Driven by Interface Status Down (Cisco Backup Interface)
   Dialing Driven by Routing Table Entry Missing (Cisco Dialer Watch)
   Dialing Driven by Best Route to Destination (Dial-on-Demand)
   Choosing a Dial Backup Approach
Basic ISDN Dial Backup.
   Configuration Example: Dial Backup Using Cisco Backup Interface
   Configuration Example: Cisco Backup Interface Options
   Configuration Example: Dial Backup Using Dialer Watch
   Configuration Example: Dial Backup Using Dial-on-Demand

5. Advanced Dial Backup.
Dial Backup over Asynchronous Links.
   General Difference Between POTS and ISDN
   Cisco Implementation Considerations
   Configuration Example: External Modem on the AUX Port to a PRI with Digital Modems
Combining Channel for Higher Bandwidth.
   Multilink PPP
   Configuration Example: Multilink PPP with Asynchronous Modems
   Configuration Example: BRI to BRI Multilink PPP
   Load Sharing of Equal Cost Links
   Configuration Example: 2.5 Mbps ISDN Dial Backup
Supporting IPX along with IP.
   Configuration Example: IPX Dial Backup with IPXRIP
   Configuration Example: Dial Backup with OSPF for IP and IPXRIP for IPX
BGP-Driven Dial Backup.
   Configuration Example: Dial Backup for Dial Backup (IP and IPX)
   Configuration Example: Dial Backup for Multiple Primary Links

6. Multiple Routers at a Single Site.
Protecting LAN Users from Router Loss.
   Passive RIP
   Configuration Example: Passive RIP
   Proxy ARP
   IRDP, BootP, and DHCP
   VRRP and Cisco HSRP
   Configuration Example: Simple HSRP
   Configuration Example: Load Balancing with HSRP
   Configuration Example: Meeting Special Needs with HSRP
Dial Backup by One Router for a Line on Another.
   Configuration Example: Using Dialer Watch
   Configuration Example: Using Dial-on-Demand
   Configuration Example: Using BGP-Driven Dial-on-Demand
Protecting against LAN Segmentation.
   A Workaround for IP Networks
   Configuration Example: Single LAN with Two WAN Routers
   Configuration Example: Putting It All Together

7. Hub and Spokes Topology.
Routing Protocols for Hubs with Many Spokes.
   Routing Requirements for Hub Routers versus Spoke Routers
   Configuration Example: EIGRP Everywhere
   Configuration Example: OSPF Hub with RIPv2 Spokes
   Configuration Example: Dual Connections to Every Spoke
Dial Backup with Multiple Dial Access Servers.
   Eliminating Dial Access Server Dependency
   Configuration Example: Single Spoke, Two Core Targets, IP Only
   Configuration Example: Single Spoke, Two Core Targets, IP and IPX
   Configuration Example: Multiple ISDN and Async Targets
Using More than One Hub.
   Dealing with Loss of Hub Network Connectivity
   Configuration Example: Hub Dialing Around Loss of Hub Internal Connectivity

8. Connecting to Service Providers.
Multiple Links to an External Network.
   Configuration Example: Redundant Static Routes
   Configuration Example: Static Routes Driven by an IGP
   Configuration Example: Primary Path with Dial Backup
   Configuration Example: Using NAT to Route Return Traffic
Connecting to an Internet Service Provider.
   Challenges for Higher Internet Availability
   Configuration Example: Static Routing with Two Internet Links
   Routing with BGP
   Configuration Example: Using BGP for Link Fault Detection
   Configuration Example: BGP to ISP with Dynamic Internal Routing
   Configuration Example: Detecting ISP Failures with BGP
Connecting through Multiple Internet Service Providers.
   Configuration Example: Using NAT for an Alternate ISP Path
   Interdomain Routing with BGP
   Configuration Example: Multihomed Organization Speaking BGP
   Configuration Example: Multihomed BGP with Redundant Routers

9. Connecting through Firewalls.
Firewalls Concepts.
   How Firewalls Are Used
   State Sensitive versus Stateless Firewalls
   Router Mode versus End-System Mode Firewalls
Firewalls in a Fully Redundant Network.
   Configuration Example: Redundant Network with Firewalls
Automating Failover in Redundant Firewall Networks.
   Router Controlled Failover Using Mirrored Router Mode Firewalls
   Configuration Example: Automatic Failover with Router Mode Firewalls
   Router Controlled Failover of Alternate End-System Mode Firewalls
   Configuration Example: Automatic Failover with End-System Mode Firewalls

10. IBM Mainframe Connections.
Data Link Switching (DLSw) Protocol.
   How DLSw Works
   Configuration Example: Redundant DLSw between Token Rings
DLSw for Ethernet Attached Devices.
   Redundant DLSw versus Transparent Bridging
   Redundant DLSw without Source Route Bridging
   Configuration Example: Redundant Ethernet Peers using Backup Peers
   Configuration Example: Backup Peers on a Segmentable LAN
   Configuration Example: Redundant Peers using DLSw+ Ethernet Redundancy
DLSw through Firewalls.
   DLSw Backup Peers for Firewall Redundancy
   Configuration Example: Token Ring DLSw through Dual NAT Firewalls
   Configuration Example: Ethernet DLSw through Dual NAT Firewalls

11. Disaster Recovery Considerations.
Disaster Recovery Requirements.
   Disaster Recovery Planning
   Disaster Recovery Approaches
Traditional Approaches.
Hot Backup Facility Approaches.
   One IP Address for Two Sites
   Configuration Example: Backup Server and Network Address Translation
   Configuration Example: Backup Server using NAT and Host Routing
   Domain Name System Operation
   Domain Name System Redirection
Load Sharing Approaches.
   Load Sharing Using Domain Name System Redirection
   HTTP Redirection

12. Management Considerations.
Network Monitoring.
   Network Management with SNMP
   Monitoring Functions that Do Not Report Status
   Routine Testing of Backup Facilities
Configuration Management.
   Configuration Documentation Matches Reality
   Configuration is Understood
   Managing Configuration Changes
   Recovering from Emergency Repair Efforts
Total Quality Control.
   Design Verification
   Error Analysis




| Home Page | Company Profile | Capabilities | Coming Events | Case Studies | White Papers | Book

Copyright 1999-2000 © Networking Unlimited Inc. All rights reserved.