- •9 Ethernet
- •9.0 Chapter Introduction
- •9.0.1 Chapter Introduction Page 1:
- •9.1 Overview of Ethernet
- •9.1.1 Ethernet - Standards and Implementation Page 1:
- •Ieee Standards
- •9.1.2 Ethernet - Layer 1 and Layer 2 Page 1:
- •9.1.3 Logical Link Control - Connecting to the Upper Layers Page 1:
- •9.2 Ethernet - Communication through the lan
- •9.2.1 Historic Ethernet Page 1:
- •9.1.5 Physical Implementations of Ethernet Page 1:
- •9.2.2 Ethernet Collision Management Page 1:
- •9.2.3 Moving to 1Gbps and Beyond Page 1:
- •9.3 The Ethernet Frame
- •9.3.1 The Frame - Encapsulating the Packet Page 1:
- •9.3.2 The Ethernet mac Address Page 1:
- •9.3.3 Hexadecimal Numbering and Addressing Page 1:
- •Viewing the mac
- •9.3.4 Another Layer of Addressing Page 1:
- •9.3.5 Ethernet Unicast, Multicast & Broadcast Page 1:
- •9.4 Ethernet Media Access Control
- •9.4.1 Media Access Control in Ethernet Page 1:
- •9.4.2 Csma/cd - The Process Page 1:
- •9.4.3 Ethernet Timing Page 1:
- •9.4.4 Interframe Spacing and Backoff Page 1:
- •Interframe Spacing
- •9.5 Ethernet Physical Layer
- •9.5.1 Overview of Ethernet Physical Layer Page 1:
- •9.5.2 10 And 100 Mbps Ethernet Page 1:
- •10 Mbps Ethernet - 10base-t
- •100 Mbps - Fast Ethernet
- •100Base-tx
- •100Base-fx
- •9.5.3 1000 Mbps Ethernet Page 1:
- •1000 Mbps - Gigabit Ethernet
- •1000Base-t Ethernet
- •1000Base-sx and 1000base-lx Ethernet Using Fiber-Optics
- •9.5.4 Ethernet - Future Options Page 1:
- •9.6 Hubs and Switches
- •9.6.1 Legacy Ethernet - Using Hubs Page 1:
- •9.6.2 Ethernet - Using Switches Page 1:
- •9.6.3 Switches - Selective Forwarding Page 1:
- •9.6.4 Ethernet - Comparing Hubs and Switches Page 1:
- •9.7 Address Resolution Protocol (arp)
- •9.7.1 The arp Process - Mapping ip to mac Addresses Page 1:
- •9.7.2 The arp Process - Destinations outside the Local Network Page 1:
- •9.7.3 The arp Process - Removing Address Mappings Page 1:
- •9.7.4 Arp Broadcasts - Issues Page 1:
- •9.8 Chapter Labs
- •9.9 Chapter Summary
- •9.9.1 Summary and Review Page 1:
- •9.10 Chapter Quiz
- •9.10.1 Chapter Quiz Page 1:
9.6.2 Ethernet - Using Switches Page 1:
In the last few years, switches have quickly become a fundamental part of most networks. Switches allow the segmentation of the LAN into separate collision domains. Each port of the switch represents a separate collision domain and provides the full media bandwidth to the node or nodes connected on that port. With fewer nodes in each collision domain, there is an increase in the average bandwidth available to each node, and collisions are reduced.
A LAN may have a centralized switch connecting to hubs that still provide the connectivity to nodes. Or, a LAN may have all nodes connected directly to a switch. Theses topologies are shown in the figure.
In a LAN where a hub is connected to a switch port, there is still shared bandwidth, which may result in collisions within the shared environment of the hub. However, the switch will isolate the segment and limit collisions to traffic between the hub's ports.
Page 2:
Nodes are Connected Directly
In a LAN where all nodes are connected directly to the switch, the throughput of the network increases dramatically. The three primary reasons for this increase are:
-
Dedicated bandwidth to each port
-
Collision-free environment
-
Full-duplex operation
These physical star topologies are essentially point to point links.
Click the performance factors in the figure.
Dedicated Bandwidth
Each node has the full media bandwidth available in the connection between the node and the switch. Because a hub replicates the signals it receives and sends them to all other ports, classic Ethernet hubs form a logical bus. This means that all the nodes have to share the same bandwidth of this bus. With switches, each device effectively has a dedicated point-to-point connection between the device and the switch, without media contention.
As an example, compare two 100 Mbps LANs, each with 10 nodes. In network segment A, the 10 nodes are connected to a hub. Each node shares the available 100 Mbps bandwidth. This provides an average of 10 Mbps to each node. In network segment B, the 10 nodes are connected to a switch. In this segment, all 10 nodes have the full 100 Mbps bandwidth available to them.
Even in this small network example, the increase in bandwidth is significant. As the number of nodes increases, the discrepancy between the available bandwidth in the two implementations increases significantly.
Collision-Free Environment
A dedicated point-to-point connection to a switch also removes any media contention between devices, allowing a node to operate with few or no collisions. In a moderately-sized classic Ethernet network using hubs, approximately 40% to 50% of the bandwidth is consumed by collision recovery. In a switched Ethernet network - where there are virtually no collisions - the overhead devoted to collision recovery is virtually eliminated. This provides the switched network with significantly better throughput rates.
Full-Duplex Operation
Switching also allows a network to operate as a full-duplex Ethernet environment. Before switching existed, Ethernet was half-duplex only. This meant that at any given time, a node could either transmit or receive. With full-duplex enabled in a switched Ethernet network, the devices connected directly to the switch ports can transmit and receive simultaneously, at the full media bandwidth.
The connection between the device and the switch is collision-free. This arrangement effectively doubles the transmission rate when compared to half-duplex. For example, if the speed of the network is 100 Mbps, each node can transmit a frame at 100 Mbps and, at the same time, receive a frame at 100 Mbps.
Using Switches Instead of Hubs
Most modern Ethernet use switches to the end devices and operate full duplex. Because switches provide so much greater throughput than hubs and increase performance so dramatically, it is fair to ask: why not use switches in every Ethernet LAN? There are three reasons why hubs are still being used:
-
Availability - LAN switches were not developed until the early 1990s and were not readily available until the mid 1990s. Early Ethernet networks used UTP hubs and many of them remain in operation to this day.
-
Economics - Initially, switches were rather expensive. As the price of switches has dropped, the use of hubs has decreased and cost is becoming less of a factor in deployment decisions.
-
Requirements - The early LAN networks were simple networks designed to exchange files and share printers. For many locations, the early networks have evolved into the converged networks of today, resulting in a substantial need for increased bandwidth available to individual users. In some circumstances, however, a shared media hub will still suffice and these products remain on the market.
The next section explores the basic operation of switches and how a switch achieves the enhanced performance upon which our networks now depend. A later course will present more details and additional technologies related to switching.
9.6.2 - Ethernet - Using Switches The diagram depicts features of switch-based LAN's. Initially, six hosts and a router are connected to a central switch. Then another switch is connected to the first one with six more hosts attached. Features include the following: -Dedicated Bandwidth: Each host accesses full network bandwidth. -Collision-Free: Network free of collisions, multiple collision domains. -Full-Duplex: Full-duplex allows communication in both directions at the same time.
Page 3:
In this activity, we provide a model for comparing the collisions found in hub-based networks with the collision-free behavior of switches.
Click the Packet Tracer icon for more details.
9.6.2 - Ethernet - Using Switches Link to Packet Tracer Exploration: From Hubs to Switches In this activity, we provide a model for comparing the collisions found in hub-based networks with the collision-free behavior of switches.
