Network Devices

Bridge

A bridge is an older version of a switch. It has the same basic functionality of a switch, but it has fewer ports and is software based, rather than hardware based (like a modern switch).

Broadcast domain

A broadcast domain is a network segment on which broadcasts occur. Microsegmentation will not stop broadcasts. Because switches flood broadcasts out all ports by default, a single switch, or any number of switches connected together, comprise a single broadcast domain. Routers block broadcasts by default, so they become the point at which the broadcast domain ends. If a router has two Ethernet interfaces, the network has two broadcast domains, one on either side of the router.

Collision domain

A collision domain is a network segment in which a collision can happen. In a collision domain, nodes contend for access to the same physical medium. This occurs on a logical bus, where the transmission of a single node is heard by all nodes. A single hub creates a single collision domain, because all nodes hear all transmissions from all other nodes. Likewise, a coax bus (which is both a physical and logical bus) is a single collision domain because the transmission of one node fills the entire medium, potentially colliding with other nodes that try to transmit.

Content switch

A content switch supports load balancing among server groups and firewalls, and web cache and application redirection, in addition to other server management functions. Content switches are often referred to as 4-7 switches as they primarily work on Layers 4 and 7 of the OSI model. They make intelligent decisions about data by analyzing data packets in real time, and understanding the criticality and type of the request. Content switching supports load balancing for servers by directing traffic to assigned server groups that perform the function. This increases the response time for requests on the network. Although complex to implement, a content switch can perform many critical functions on a network and increase throughput.

Firewalls

A firewall is a software program or a hardware device or a combination of both that protects a device or network from unauthorized data by blocking unsolicited traffic. Firewalls generally are configured to block suspicious or unsolicited incoming traffic, but allow incoming traffic sent as a response to requests from internal hosts.

Hub

A hub, or multiport repeater, is a networking device used to connect the nodes in a physical star topology network into a logical bus topology. A hub contains multiple ports to which the devices can be connected. When a data packet from the transmitting device arrives at a port, it is copied and transmitted to all other ports so that all other nodes receive the packets. However, only the node to which it is addressed reads and processes the data while all other nodes ignore it.

Multilayer switch

A multilayer switch performs both routing and switching. Also referred to as a Layer 3 switch or a Layer 2-3 switch, it can perform only limited routing functions and supports only Ethernet connections. Multilayer switches support the configuration of virtual local area networks (VLANs), which are discussed later in this topic.

Disabled

A port that has been manually disabled.

Repeater

A repeater is a device that regenerates a signal to improve signal strength over transmission distances. By using repeaters, you can exceed the normal limitations on segment lengths imposed by various networking technologies.

Routers

A router is a networking device that connects multiple networks. Operating at Layer 3 of the OSI model, it makes forwarding decisions based on Layer 3 addresses, such as Internet Protocol (IP) addresses. When a packet comes in one of the router's interfaces, the router reads the destination IP address and forwards the packet out the appropriate interface. If necessary, it will strip off the packet's Layer 2 encapsulation and replace it with encapsulation that is appropriate for the outgoing transmission medium; for example, replacing Ethernet with Point-to-Point Protocol (PPP).

Switches

A switch is a network device that acts as a common connecting point for various nodes or segments. Working at Layer 2 of the OSI model, switches make forwarding decisions based on Layer 2 (MAC) addresses. A switch listens for the MAC addresses of all the nodes plugged into it, and builds a table in memory that maps each MAC address with its associated port. When an Ethernet frame comes into the switch, the switch reads the destination MAC address from the header and consults its table to determine which port to repeat the frame out of. In this way, switches can keep conversations limited to only the nodes that are involved. Thus, a 24-port switch can have 12 pairs of conversations going on at the same time.

VLANs

A virtual LAN (VLAN) is a logical grouping of ports on the switch. An administrator determines which ports are grouped together. Nodes that plug into those ports can communicate only with nodes that belong to the same VLAN. They cannot communicate with nodes that belong to other VLANs. This effectively divides a physical switch into multiple, smaller logical switches. When a VLAN port group is extended to another device, then tagging is used. Since communications between ports on two different switches travel via the uplink ports of each switch involved, every VLAN containing such ports must also contain the uplink port of each switch involved, and these ports must be tagged. This also applies to the default VLAN. The default VLAN is the default VLAN on a switch that is used unless another VLAN is created and specified. On a Cisco switch, that default VLAN is VLAN1, which cannot be removed or renamed. If the VLAN port group exists on only a single device, then those ports would be untagged. A Native VLAN is a VLAN that handles traffic that is not tagged.

Analog Modems

An analog modem is a device that modulates signals to encode digital information and demodulates signals to decode the transmitted information. A common type of modem is one that takes the digital data of a device and turns it into modulated electrical signal for transmission over telephone lines, which is then demodulated by another modem at the receiver side to recover the digital data. Though modems are not used much anymore, they can be used in locations where you have no other options for connections.

Listening

BPDUs are processed by the switch. It does not forward frames. It waits for information to determine if it should return to the blocked state. It does not update the MAC tables.

IEEE 802.1q

Because trunk links carry all VLAN traffic, there must be some mechanism for identifying which frame belongs to which VLAN as it moves from switch to switch. IEEE 802.1q is the most commonly used trunk link protocol to address this issue. 802.1q inserts a special tag in the Ethernet header identifying the VLAN for that frame. The switch at the other end of the trunk link will read that tag and forward the frame to the appropriate VLAN. A competitor to 802.1q is Cisco's inter-switch link (ISL). This protocol encapsulates an Ethernet frame into a proprietary format that identifies the VLAN. ISL is an older protocol. Most Cisco switches now use 802.1q.

Core routers

Core routers are located at the center of network backbones. They are used to connect multiple distribution routers located in different buildings to the backbone.

Forwarding

Data is sent and received on a port in this state. BPDUs are monitored to determine if the port should return to the blocked state.

Cut-through switching mode

In cut-through switching, the switch forwards a data packet as soon as it receives it; no error checking or processing of the packet is performed. The switch performs the address table lookup immediately upon receiving the Destination Address field in the packet header. The first bits in a packet are sent out of the outbound port on a switch immediately after it receives the bits. The switch does not discard packets that are corrupt and fail error checking.

Fragment-free switching mode

In fragment-free switching, the switch scans the first 64 bytes of each packet for evidence of damage by a collision. If no damage is found, it forwards the packet; otherwise, it discards it. Fragment-free switching reduces network congestion by discarding fragments. It is similar to the cut-through switching method, but the switch waits to receive 64 bytes before it forwards the first bytes of the outgoing packet.

Store-and- forward switching mode

In store-and-forward switching, the switch calculates the cyclic redundancy check (CRC) value for the packet's data and compares it to the value included in the packet. If they match, the packet is forwarded. Otherwise, it is discarded. This is the slowest type of switching mode. The switch receives the entire frame before the first bit of the frame is forwarded. This allows the switch to inspect the frame check sequence (FCS) before forwarding the frame. FCS performs error checking on the trailer of an Ethernet frame.

Learning

Learns source addresses from received frames. Addresses are added to the switching database. MAC address table is updated. No frames are forwarded.

Network Controllers

Network controllers support large-scale interactive networks and communication between set-tops and application servers. A set-top is an information appliance device that contains a television-tuner input and displays output to a television set. They are used in cable television, satellite television, and over-the-air television systems. Network controllers are used in digital cable networks and enable such services as video-on-demand (VOD), catalog shopping, web browsing, and email. Network controllers act as a gateway between the IP network, which connects application servers, and the digital cable network, which connects set-tops.

Port Mirroring

Port mirroring is the practice of duplicating all traffic on one port in a switch to a second port, effectively sending a copy of all the data to the node connected to the second port. This is known as local port mirroring. Remote port mirroring implements port mirroring between multiple devices. In this case, the source port is on one device and the destination port is located on a different device. Port mirroring is useful as a diagnostic tool when you need to monitor all traffic going to a particular port or node with minimal impact on the network performance.

Distribution routers

Routers that collect data from multiple access routers and redistribute them to an enterprise location such as a company's headquarters. The routing capabilities of a distribution router are greater than those of access routers.

Access routers

Routers used in small office/home office (SOHO) networks. They are located at customer sites and are inexpensive.

RSTP

STP is an old protocol. Its full process to determine redundancy actually takes 50 seconds, which is considered too long by modern standards. It has since been replaced by the Rapid Spanning Tree Protocol (RSTP), IEEE 802.1w, which takes only 20 seconds to identify and rectify loops. There is also Shortest Path Bridging (SPB), which is another replacement for STP. It is intended to simplify the creation and configuration of networks, while enabling multipath routing.

STP (IEEE 802.1d)

The Spanning Tree Protocol (STP) is a Layer 2 protocol used to prevent switching loops. Whenever there are redundant paths between switches, where either two switches are connected using two different links or a ring of switches is connected to each other, a switching loop will occur. Because switches, by their nature, flood broadcasts and multicasts out all ports, the first Address Resolution Protocol (ARP) frame sent by a client trying to find a neighbor or a router will cause a Layer 2 broadcast storm. The ARP broadcast will go down one link to the next switch, which will send the broadcast back up the redundant link. This feedback loop will continue indefinitely until there is manual intervention by an administrator. It will cause network utilization to go to near maximum capacity, and the CPU utilization of the switches to jump to 80 percent. This makes the switched segment effectively unusable until the broadcast storm stops.

The TCP/IP Model

The TCP/IP model is a four-layer model developed by the United States Department of Defense. To some extent, it is similar to the OSI model. The TCP/IP model was developed to allow the addition of new technologies and create a more flexible architecture, which can easily allow the modification of existing protocols. This architecture later became known as the TCP/IP model after two of its most important protocols: TCP and IP.

VTP

The VLAN Trunking Protocol (VTP) is the messaging protocol that switches use to update each other's VLAN databases. Developed by Cisco, it allows switches to quickly advertise to each other when a VLAN is created or deleted. This saves an administrator some manual labor. If the administrator wishes to extend a VLAN across several switches, he or she would have to manually configure each switch with the same VLANs. With VTP, this is done automatically. There are three VTP modes that a switch can use: server, client, and transparent. • Server mode: This is the default mode for VTP on a switch. In the server mode, a switch can modify VLANs. This information is then transmitted to all the other switches that are configured to the same group using VTP. • Client mode: In the client mode, a switch cannot modify VLANs but will receive configuration information from other switches. • Transparent mode: In the transparent mode, a switch receives configuration messages from other switches but does not process them. Configuration changes to the VLAN are not transmitted to other switches in the group.

Gateway

The word gateway is a generic term for any device or software that translates one network protocol to another. Working at OSI Layer 3 or above, gateways connect incompatible systems by taking an incoming packet, stripping off the lower-level encapsulation of the original protocol, and re- encapsulating the packet with a new protocol.

Trunking

Trunk links can be combined to increase bandwidth and reliability in a process called trunking.. This is also known as link aggregation, port bonding, port teaming, EtherChannel, and NIC bonding, among other names. Although a variety of manufacturer-implemented techniques exist, IEEE 802.1AX-2008 defines a standard for link aggregation. Within the IEEE specification, the Link Aggregation Control Protocol (LACP) provides a method to control the bundling of several physical ports together to form a single logical channel. LACP allows a network device to negotiate an automatic bundling of links by sending LACP packets to the peer. The primary purpose of link aggregation is to allow redundant links to combine their bandwidth together without causing spanning tree loops. Link aggregation is typically implemented between switches, although it can be implemented between a node and a switch. Linking two 1-Gbps ports on a server to two 1-Gbps ports on a switch can result in 2 Gbps aggregate throughput. Depending on the implementation, this can result in a redundant connection in case one of the cables or ports fails. However, this still leaves the possibility of the entire switch failing, so some hardware vendors provide proprietary methods for trunking ports across two physically separate switches. Trunking can be used to connect a variety of network hardware, including switch-to-switch, server-to-switch, server-to-server, or switch-to-router.

Unmanaged vs. Managed Switches

Unmanaged switches are devices that perform switching without user intervention. In other words, the functions of an unmanaged switch cannot be controlled. On the other hand, a managed switch provides complete control over how the device functions. It is typically assigned its own IP address and can be managed either directly through the use of a console cable, a Telnet or as a Secure Shell (SSH) connection, or a web browser. Managed switches allow administrators to create virtual local area networks (VLANs) within the network.

Blocking

User data is not sent or received. BPDU data is received, but the port only goes into another state if other links are unavailable and the spanning tree algorithm determines that the port should change to the forwarding state.

VLAN Pooling

VLAN pooling is a mechanism whereby WAPs can choose among several different available VLANs to assign to incoming client connections. This strategy distributes and load-balances wireless client traffic among multiple VLANs so that no single network segment is overwhelmed by too many wireless client connections.

passive hub

has its ports wired together physically. It connects devices plugged into it without the use of power. Acting like a patch panel, it merely makes the electrical connection without repeating or transmitting any frames. A token ring MSAU is an example of a passive hub.

active hub

is a true multiport repeater. It receives incoming frames and retransmits those frames out all ports. An Ethernet hub is an example of an active hub.