Previous Table of Contents Next


Hubs and Switches

Hubs and switches reflect the trend toward star, and away from bus, configurations. Hubs can be either active or passive. Passive hubs act simply as cable-connecting devices, while active hubs also serve as signal repeaters [9-2]. The first generation of hubs (1984) acted as LAN concentrators and repeaters, with a single internal collapsed backbone bus for connecting like LANs. The second generation accommodated multiple LAN architectures (Ethernet and Token-Passing Ring) over separate ports; rudimentary network management and configuration capabilities were often included as well [9-13].

A collapsed backbone is a fairly simple concept, and one worth exploring in some detail. LANs traditionally work on the basis of a common electrical bus (shared cable medium) to which each device or group of devices is directly connected. While this approach works effectively, it requires that the cable be deployed through the entire workplace. The traditional coax medium is expensive to acquire and deploy as is fiber optic cable. Additionally, the cable is susceptible to physical damage unless it is protected by conduits, armoring or some other means, all of which involve additional cost. Alternatively, the high-speed backbone bus can be collapsed and placed within a hub, to which the devices are connected by means of a less expensive medium such as UTP. In this manner, the backbone bus is protected and the cost of cabling is much reduced. Should an individual device (workstation, NIC or MAU) create difficulty, it can be isolated by disabling the port by which is attached to the hub. Should a UTP cable suffer damage, the connection can be similarly disabled and the problem corrected at relatively low cost. While the UTP cable is inherently less capable than coax or fiber, equivalent bandwidth can be provided as long as the distances between the devices and the hub are within tolerable limits. The above is an accurate description of a 10Base-T configuration.

Intelligent Hubs, the third generation, provide multiple buses for multiple LANs, either of the same or disparate architectures (Figure 9.8). They support multiple media (coax, UTP, and fiber), multiple speeds, and multiple LAN protocols. As addressable devices, they can be centrally managed via SNMP (Simple Network Management Protocol) or another appropriate network management protocol. Intelligent hubs also provide bridging and basic routing capabilities [9-13]. Intelligent hubs with basic routing capabilities are also known as LAN switches; they can read the target address of the packet and forward it to the appropriate LAN segment, filtering it in the process in order to avoid its contributing to congestion on other LAN segments. The LAN switch also offers the advantage of a high-capacity, multibus switching matrix that can provide full LAN bandwidth to multiple, simultaneous communications on a point-to-point basis. For instance, one workstation can access another over a 10-Mbps connection, while another has a full 10-Mbps connection to a database server, and while still another is passing a file to a print server at high speed [9-16].


Figure 9.8  Intelligent hub configuration.

Regardless of the generation, the hubs serve, at minimum, as central points of interconnection for LAN-attached devices. Additionally, hubs serve as concentrators of LAN traffic and as repeaters, with multiple hubs interconnected through high-speed media [9-14]. A number of hub manufacturers offer stackable hubs, which offer the advantage of scalability; in other words, the hubs can be stacked and interconnected in order to increase port and traffic capacity. While this is an inexpensive approach to increasing capacity, there is a tradeoff in terms of the lack of high-speed interhub busses in a stack configuration [9-17].

Routers

Routers are highly intelligent devices that support connectivity between both like and disparate LANs, as shown in Figure 9.9. Additionally, routers provide connectivity to MANs (SMDS) and WANs (X.25, Frame Relay and ATM). Routers are protocol-sensitive, typically supporting multiple protocols, and large and varying packet sizes such as might be involved in supporting both Ethernet and Token Ring. Routers operate at the bottom 3 layers of the OSI model, using the Physical Layer, Link Layer and Network Layer to provide connectivity, addressing and switching [9-14].


Figure 9.9  Router network.

Server-based routers also are available in the form of router software embedded in a server. As they perform more slowly than their hardware relatives they generally are not considered up to the task of serving an enterprise-wide application. They do have application, however, in support of smaller, remote sites [9-18].

In addition to supporting filtering and encapsulation, routers route traffic based on a high level of intelligence that allows them to consider the network as a whole. Hubs and bridges, on the other hand, simply view the network on a link-by-link basis. Routing considerations might include destination address, packet priority level, least-cost route, minimum route delay, minimum route distance, and route congestion level. Routers are also self-learning, as they can communicate their existence to other devices and can learn of the existence of new routers, nodes and LAN segments. Routers constantly monitor the condition of the network, as a whole, in order to dynamically adapt to changes in the condition of the network. Routers are multiport devices with high-speed backbones which can be on the order of 1 Gbps+. Additionally, they typically provide some level of redundancy in order that they are less susceptible to catastrophic failure [9-13].

Interesting recent developments in the world of routers include those of alternate routing and inverse multiplexing. As noted by Charles Darling [9-19], what dedicated-WAN links lack in cost efficiency they make up for in terms of lack of reliability (backhoe-fade). Select manufacturers now offer routers that sense network failure and re-establish the connection via an alternate connection. While a backup ISDN link may be painfully slow compared with a T1 connection, Darling suggests that slow is better than stop. Additionally, some routers offer inverse multiplexing capabilities over ISDN and T-carrier facilities.

Router protocols include both bridging and routing protocols. These protocols fall into three categories: inter-router protocols, serial line protocols, and protocol stack routing and bridging.


Previous Table of Contents Next