Chapter 6 presents the various approaches to achieving high-availability network designs, including trade-offs and recommendations. We describe various techniques, provide some detailed configuration examples of the different ways to connect servers to the edge switches, and touch on how data center switches can be configured for increased availability, as shown in FIGURE 1-8. The material in Chapter 6 is based on actual customer experiences and has proven to be quite valuable to the network architect. We describe the following Layer 2 approaches:
Trunking – NIC, server side
Trunking – Switch side, including Distributed Multi-link Trunking (DMLT)
Spanning Tree Protocol (STP)
The following Layer 3 strategies are described:
Virtual Router Redundancy Protocol (VRRP) – default router redundancy mechanisms
IP Multipathing (IPMP) – NIC redundancy
Open Shortest Path First (OSPF) and Routing Information Protocol (RIP) – data center routing protocol availability features.
The advantages and disadvantages will be described, along with suggestions on which approach makes sense for which situation.