1.2 Key Characteristics of Scalable Internetworks
1.2.2 Making the network reliable and available
The internetwork should be reliable and available at all layers, but most critically at the core layer. Recall that core routers must be reliable because they carry information about all the routes in an internetwork. If one of these routers goes down, it affects routing on a large scale.

Core routers are reliable when they can accommodate failures by rerouting traffic and respond quickly to changes in the network topology. Some protocols that enhance network reliability and availability that the Cisco IOS software supports are scalable routing protocols, tunnels, and dial backup.

Scalable Protocols

Scalable protocols include Open Shortest Path First (OSPF) and Enhanced Interior Gateway Routing Protocol (EIGRP). These protocols provide the following features:

Reachability --- Scalable internetworks, even those using a hierarchical design, can have a large number of reachable networks or subnetworks. These networks can be subject to reachability problems due to metric limitations of distance vector routing protocols. Scalable routing protocols, such as OSPF, NetWare Link Services Protocol (NLSP), and EIGRP, use metrics that expand the reachability potential for routing updates because they use cost, rather than hop count, as a metric.

Fast convergence time --- Convergence time is defined as the amount of time required to propagate new route information from one end of the internetwork to the other end of the internetwork. Scalable protocols can converge quickly because the router can detect failure rapidly and because each router maintains a network topology map. Routers also forward network changes quickly to all routers in the network topology.

Congestion control --- Scalable routing protocols generally add less traffic overhead to the network for carrying routing information by providing summarizations of network information.

Alternate Paths

Many internetwork backbones carry mission-critical information. Organizations running such backbones are usually interested in protecting the integrity of this information at virtually any cost. Routers must offer sufficient reliability so that they are not the weak link in the internetwork chain. The key is to provide alternative paths that can come on line whenever link failures occur along active networks.

End-to-end reliability is not ensured by simply making the backbone fault tolerant. If communication on a local segment within any building is disrupted for any reason, that information will not reach the backbone. End-to-end reliability is possible only when redundancy is employed throughout the internetwork. Because this redundancy is usually cost-prohibitive, most companies prefer to employ redundant paths on only those segments that carry mission-critical information.

What does it take to make the backbone reliable? Routers hold the key to reliable internetworking. Depending on the definition of reliability, this can mean duplicating every major system on each router and possibly every component. However, hardware component duplication is not the entire solution because extra circuitry is necessary to link the duplicate components to allow them to communicate. This solution is usually very expensive, but more importantly, it does not completely address the problem. Even assuming all routers in your network are completely reliable systems, link problems between nodes within a backbone can still defeat a redundant hardware solution.

To really address the problem of network reliability, links must be redundant. Further, it is not enough to simply duplicate all links. Dual links must terminate at multiple routers unless all backbone routers are completely fault tolerant (no single points of failure). Otherwise, backbone routers that are not fault tolerant become single points of failure. The inevitable conclusion is that a completely redundant router is not the most effective solution to the reliability problem because it is expensive and still does not address link reliability.

Scalable protocols, such as EIGRP and OSPF, enable a router to maintain a map of the entire network topology, so when a failure is detected the router can reroute traffic by looking at the network topology and finding another path. EIGRP is also a feasible solution because it keeps a record of alternate routes in case the preferred route goes away.

Load Balancing

Load balancing is the easiest way to add bandwidth in a network with multiple links. Routers provide built-in load balancing for multiple links and paths. You can use up to four paths to a destination network. In some cases, the paths do not need to be of equal cost.

Within IP, routers provide load balancing on both a per-packet and a per-destination basis. For per-destination load balancing, each router uses its route cache to determine the output interface. If IGRP or EIGRP routing is used, unequal-cost load balancing is possible. The router uses metrics to determine which paths the packets will take; the user can adjust the amount of load balancing.

Because scalable protocols have a map of the entire network topology, and because of the way in which they maintain their routing tables, they are able to transport data simultaneously across multiple paths to a given location, as shown in the figure to the left.

Tunnels

Software tunnels can provide communication across WAN links into network areas that were previously unreachable. Tunnels allow you to configure a point-to-point link between two discontiguous networks running a given protocol, such as IPX, without configuring the entire cloud for IPX. Not only does this provide availability, but it also eliminates the overhead associated with running an additional routing protocol over the link. In addition, tunneling provides a means for encapsulating packets inside a routable protocol via virtual interfaces.

Dial Backup

On WAN connections, you can configure backup links when you need to perform the following tasks:

  • Make the primary WAN connection more reliable by configuring one or more on backup connections.
  • Increase availability by configuring the backup connections to be used when a primary connection is experiencing congestion.