Cisco LAN Design & Switched Networks

Cisco LAN Design & Switched Networks

In this article you will learn about the Cisco LAN Design & Switched Networks. Different network models and hierarchies are shared here. After reading this article you will have a clear understanding of how to design a Cisco Lan or switched network.
The digital world is changing. The ability to access the Internet and the corporate network is not the same and now, employees can access resources from anywhere in the world, and the information must be available at any time and on any device.
To support collaboration, commercial networks employ convergent solutions through voice systems, IP phones, voice gateways, video support and video conferencing. This physical network must be properly designed and implemented to allow the reliable handling of the various types of information it must carry. To allow the administration of this complex environment, a structured design is required.

Cisco LAN Network Design

One of the most recent developments in network design is Cisco Borderless Networks .

Cisco Borderless Network is a network architecture that combines innovation and design. It allows organizations to support a borderless network that can connect with anyone, anywhere, at any time, on any device, in a secure, reliable and inconvenient way. This architecture is designed to meet commercial and IT challenges, such as the admission of converging networks and changing work patterns.
With this architecture, the borderless network is built on a hierarchical hardware infrastructure that is scalable and recoverable, as shown in Image

Hierarchy in Switched Networks without borders

The creation of a borderless switched network requires the use of strong network design principles to ensure maximum availability, flexibility, security and ease of administration. These should work according to current requirements and the services and technologies that will be required in the future. Design guidelines for switched networks without borders are based on the following principles:

• Hierarchy : facilitates the understanding of the function of each device at each level, simplifies implementation, operation and administration, and reduces error domains at each level.
• Modularity : allows the expansion of the network and the enablement of integrated services without problems and upon request.
• Recovery capacity : satisfies user expectations by keeping the network always active.
• Flexibility : allows you to share traffic load intelligently by using all network resources.

These are not independent principles. It is essential to understand how each principle fits in the context of others. The hierarchical design of a borderless switched network lays a foundation that allows network designers to superimpose the features of security, mobility and unified communication. The three and two level layer models, such as those shown in Images 2 and 3, are double-checked hierarchical design frameworks for campus networks.

3

Image 2

 The three fundamental layers within these level designs are the access, distribution and core layers . Each layer can be considered as a well-defined structured module, with specific functions and roles in the campus network. The introduction of modularity in the hierarchical design of the campus ensures even more that the campus network maintains sufficient strength and flexibility to provide fundamental network services. Modularity also allows growth and changes that occur over time.

ACCESS, DISTRIBUTION AND CORE CORE LAYERS

ACCESS LAYER

The access layer represents the perimeter of the network , where traffic from the campus network enters or exits. Traditionally, the main function of access layer switches is to provide network access to the user. The access layer switches connect to the distribution layer switches, which implement network-based technologies such as routing, quality of service and security.

DISTRIBUTION LAYER

The distribution layer interacts between the access layer and the core layer to provide many important functions, including the following:

• Add wiring closet networks on a large scale.
• Add layer 2 broadcast domains and layer 3 routing limits.
• Provide intelligent switching, routing and network access policy functions to access the rest of the network.
• Provide high availability to the end user through redundant distribution layer switches, and routes of equal cost to the core.
• Provide differentiated services to different kinds of service applications in the perimeter of the network.

CORE CORE LAYER

The core layer is the backbone of a network. It connects several layers of the campus network. The core layer functions as an aggregator for the rest of the campus blocks and links the campus with the rest of the network. The main purpose of the core layer is to provide fault isolation and high speed backbone connectivity.

ROLE OF SWITCHED NETWORKS

The role of switched networks evolved significantly in the last two decades. Not long ago, Layer 2 flat switched networks were usual, depended on Ethernet and the widespread use of hub repeaters to propagate LAN traffic through an organization

Image6

As shown in Image 6, the networks were basically switched to switched LANs in the hierarchical network. Switched LANs provide more flexibility, traffic management and additional features:

• Quality of service
• Additional security
• Compatibility with wireless networking and connectivity technology
• Compatibility with new technologies, such as IP telephony and mobility services

FORM FACTORS

In commercial networks, various types of switches are used : It is important to implement the appropriate types of switches according to the requirements of the network. The following table highlights some common business considerations that should be taken into account when selecting the switch equipment.

Commercial Considerations Table to select a switch equipment.

Factor	Description
cost	The cost of a switch depends on the quantity and speed of the interfaces, the supported functions and the expandability.
Port Density	Network switches must support an adequate number of devices in the network.
Feeding	Nowadays, it is common to power access points, IP phones and even compact switches using Ethernet power. In addition to Ethernet power considerations, some rack-based switches support redundant power supplies.
Reliability	The switch must provide continuous access to the network.
Port speed	The speed of the network connection is one of the fundamental aspects for end users.
Frame buffers	The ability of the switch to store frames is important in networks where there may be congested ports connected to servers or other areas of the network.
Scalability	In general, the number of users in a network increases over time; therefore, the switch must provide the possibility of growth.

When the switch type is selected, the network designer must choose between a fixed or a modular configuration, and between a stackable or non-stackable device. Another consideration is the thickness of the switch, expressed in number of rack units. This is important for switches that are mounted in a rack. For example, the fixed configuration switches shown in Image 3 are all 1 rack unit (1U). Often these options are called switch form factors.

FIXED CONFIGURATION SWITCHES

Fixed configuration switches do not support features or options beyond those that originally came with the switch (Image 8). The specific model determines the features and options available. For example, a 24-port fixed gigabit switch does not support additional ports. In general, there are different configuration options that vary depending on the number and type of ports included in a fixed configuration switch.

MODULAR CONFIGURATION SWITCHES

Modular configuration switches offer more flexibility in your configuration. Generally, these switches come with racks of different sizes that allow the installation of different numbers of modular line cards (Image 9). Line cards are the ones that contain the ports. The line card fits the switch frame just like the expansion cards fit on the computer. The larger the chassis, the more modules it can support. There are many different chassis sizes. A modular switch with a single 24-port line card could have an additional 24-port line card installed so that the total number of ports amounts to 48.

STACKABLE CONFIGURATION SWITCHES

Stackable configuration switches can be interconnected using a special cable that provides high bandwidth performance between t
he switches (Image 10). Cisco StackWise technology allows the interconnection of up to nine switches. The switches can be stacked on top of each other with wires that connect the switches in a daisy chain shape. Stacked switches operate effectively as a single larger switch. Stackable switches are convenient when fault tolerance and bandwidth availability are criticaland it is expensive to implement a modular switch. By cross-linking these stacked switches, the network can recover quickly if a single switch fails. Stackable switches use a special port for interconnections. Many Cisco stackable switches also support StackPower technology, which allows power to be shared between the members of the stack.

 Switched Networks

The concept of switching and frame forwarding is universal in network technology and telecommunications. In LAN, WAN and public switched telephone network (PSTN) networks, various types of switches are used. The fundamental concept of switching refers to a device that makes a decision according to two criteria:

• input port
• Destination address

The decision on how a switch forwards traffic is made in relation to the flow of that traffic. The term " input " is used to describe the location of a port through which a frame enters the device . The term " output " is used to describe the frames that leave the device from a given port .
LAN switches maintain a table they use to determine how to forward traffic through the switch.
The only intelligence that LAN switches possess is the ability to use the table to forward traffic according to the input port and the destination address of a message . With LAN switches, there is only one main switching table that describes a strict association between addresses and ports; therefore, a message with a specific destination address always leaves through the same output port, regardless of the input port through which it enters.

MAC ADDRESS TABLE OF A SWITCH

The switches use MAC addresses to direct network communications through the switch to the corresponding port to the destination. A switch consists of integrated circuits and complementary software that controls data paths through the switch. To define which port to use to transmit a frame, the switch must first know what devices exist on each port. As the switch discovers the relationship between ports and devices, it creates a table called " MAC address table " or "addressable content memory table" (CAM). CAM is a special type of memory used in high speed search applications.

LAN switches determine how to handle incoming data frames using a MAC address table. The switch generates the MAC address table by registering the MAC address of each device connected to each of the ports. The switch uses the information in the MAC address table to send frames destined for a specific device through the port that was assigned to that device.
The next two-step process is performed for each Ethernet frame that enters a switch.

STEP 1: LEARNING: EXAMINE THE SOURCE MAC ADDRESS

Each frame that enters a switch is checked for new information. This is done by examining the source MAC address of the frame and the port number through which it entered the switch:

• If the source MAC address does not exist, it is added to the table, along with the input port number.
• If the source MAC address exists, the switch updates the update timer for that entry. By default, most Ethernet switches save an entry in the table for five minutes.

Note : If the source MAC address exists in the table, but on a different port, the switch treats it as a new entry. The entry is replaced with the same MAC address, but with the most current port number.

STEP 2: FORWARDING: EXAMINE THE DESTINATION MAC ADDRESS

If the destination MAC address is a unicast address, the switch searches for a match between the destination MAC address of the frame and an entry in the MAC address table:

• If the destination MAC address is in the table, it resends the frame through the specified port.
• If the destination MAC address is not in the table, the switch forwards the frame through all ports, except the input. This is known as unicast.

Note : If the destination MAC address is broadcast or multicast, the frame is also sent over all ports, except the incoming one.

SWITCH FORWARDING METHODS

As the networks grew and companies began to experience slower network performance, Ethernet bridges (an earlier version of the switch) were added to the networks to limit the size of collision domains. In the 1990s, advances in integrated circuit technologies allowed Ethernet LAN switches to replace Ethernet bridges. These switches could transport layer 2 forwarding decisions from the software to the specific application integrated circuits (ASICs). ASICs reduce packet handling time within the device and allow the device to handle a greater number of ports without decreasing performance. This method of forwarding data frames in layer 2 was called "storage and send switching".

STORAGE AND SHIPPING SWITCHING

Switching through storage and shipping has two main characteristics that differentiate it from the cutting method: error verification and automatic buffer storage.

• Error Verification

Switches that use storage and send switching perform error verification of incoming frames. After receiving the full frame at the input port, as shown in the illustration, the switch compares the frame check sequence (FCS) value in the last field of the datagram with its own FCS calculations. FCS is an error verification process that helps ensure that the frame does not contain physical or data link errors. If the frame has no errors, the switch forwards it. Otherwise, it is discarded.

• Auto buffer storage

The buffering process of the input port used by the storage and send switches provides the flexibility to support any combination of Ethernet speeds. For example, the handling of an incoming frame that moves to a 100 Mb / s Ethernet port and that must be sent over a 1 Gb / s interface requires the use of the storage and sending method. In the event of any incompatibility of the input and output port speeds, the switch stores the entire frame in a buffer, calculates the FCS verification, forwards it to the output port buffer and then sends it.

SWITCHING BY CUTTING METHOD

An advantage of switching by cutting method is that the switch has the ability to initiate the forwarding of a frame rather than the storage and send switching. The switching by cut has two main characteristics: the fast forwarding of frames and the free switching of fragments.

• Fast frame forwarding

As indicated in the illustration, switches using the cut-off method can make a forwarding decision as soon as they find the destination MAC address of the frame in the MAC address table. The switch does not have to wait for the rest of the frame to enter the input port before making the forwarding decision.
With current MAC controllers and ASICs, switches that use the cut-off method can quickly decide if they need to examine most of the headers of a frame for additional filtering purposes.

• Fragment free

Fragment-free switching is a modified form of switching by cutting method in which the switch waits for the collision window to pass (64 bytes) before resending the frame. This means that each frame is registered in the data field to ensure that fragmentation does not occur. Fragment-free switching provides better error verification than cutting, with virtually no latency increase.

The lower latency speed of cut-off switching makes it more suitable for very demanding high-performance computing (HPC) applications that require process-to-process latencies of 10 microseconds or less.

SWITCHING DOMAINS

1. COLLISION DOMAINS

In hub-based Ethernet segments, network devices compete for the medium, because the devices must take turns during transmission. Network segments that share the same bandwidth between devices are known as collision domains . When two or more devices from the same collision domain try to communicate at the same time, a collision occurs.

If an Ethernet switch port is operating in half-duplex mode, each segment is in its own collision domain. However, Ethernet switch ports that operate in full-duplex mode eliminate collisions; Therefore, there is no collision domain. By default, the Ethernet switch ports autonegotiate the full duplex when the adjacent device can also operate in full duplex mode. If the switch port is connected to a device that works in half-duplex, such as an old hub, the switch port will operate in half-duplex mode. In the case of half-duplex, the switch port will be part of a collision domain.

2. BROADCAST DOMAINS

A series of interconnected switches forms a simple broadcast domain. Only network layer devices, such as routers, can split a layer 2 broadcast domain . Routers are used to segment broadcast domains, but they also segment a collision domain.

When a device wishes to send a layer 2 broadcast, the destination MAC address of the frame is set only to binary numbers one.
The layer 2 broadcast domain is called " MAC broadcast domain ". The MAC broadcast domain consists of all the devices on the LAN that receive broadcast frames from a host.

RELIEF FROM NETWORK CONGESTION

LAN switches have special features that make them effective in relieving congestion in a network. By default, interconnected switch ports try to establish a full-duplex link and therefore collision domains are eliminated . Each full duplex port on the switch offers full bandwidth to devices connected to that port. A full duplex connection can carry the transmitted and received signals at the same time. Full duplex connections greatly increased the performance of LAN networks and are required for Ethernet speeds of 1 Gb / s and higher.
The switches interconnect LAN segments, use a MAC address table to determine the segment to which they should send the frame and can reduce or eliminate collisions altogether. Finally, some important features of the switches that help to relieve network congestion are detailed:

• High port density : switches have high port densities; 24 and 48 port switches are often only a rack unit and operate at speeds of 100 Mb / s, 1 Gb / s and 10 Gb / s. Large business switches can support hundreds of ports.
• Large frame buffers:  The ability to store more frames received before starting to discard them is useful, especially when there may be congested ports connected to servers or other parts of the network.
• Port speed:  Depending on the cost of a switch, it is possible to support a combination of speeds. The 100 Mb / s and 1 Gb / s or 10 Gb / s ports are common (there may also be 100 Gb / s).
• Fast internal switching:  the ability of fast internal forwarding promotes high performance. The method used can be an internal bus or a high-speed shared memory, which affects the overall performance of the switch.
• Low cost per port:  switches provide high port density at a lower cost.