CCNA R&S-07-Building Ethernet LANs with Switches

©2015 Amir Jafari – www.amir-Jafari.com
Routing and Switching 200-120
7 - Building Ethernet LANs with Switches

Building Ethernet LANs with Switches
Agenda
 LAN Switching Concepts
 Design Choices in Ethernet LANs

LAN Switching Concepts
Historical Progression: Hubs, Bridges, and Switches
 10BASE-T used a centralized cabling model similar to today’s Ethernet
LANs, with each device connecting to the LAN using a UTP cable
 Instead of a LAN switch, the early 10BASE-T networks used hubs, because
LAN switches had not yet been created
10BASE-T (with a Hub)

With 10BASE-T using hubs:
 When hubs receive an electrical signal in one port , the hub repeats the signal
out all other ports
 When two or more devices send at the same time, an electrical collision
occurs, making both signals corrupt
 As a result, devices must take turns by using carrier sense multiple access
with collision detection (CSMA/CD) logic, so the devices share the (10-Mbps)
bandwidth
 Broadcasts sent by one device are heard by, and processed by, all other
devices on the LAN
 Unicast frames are heard by all other devices on the LAN

Ethernet transparent bridges helped solve this performance problem with
10BASE-T:
 Bridges separated devices into groups called collision domains
 Bridges reduced the number of collisions that occurred in the network,
because frames inside one collision domain did not collide with frames in
another collision domain
 Bridges increased bandwidth by giving each collision domain its own
separate bandwidth, with one sender at a time per collision domain

 Bridge will buffer or queue the frame until the outgoing interface can send the
frame
 Adding the bridge in Figure really creates two separate 10BASE-T networks
Bridge Creates Two Collision Domains and Two Shared
Ethernets

 LAN switches perform the same basic core functions as bridges, but at much
faster speeds and with many enhanced features
 Like bridges, switches segment a LAN into separate collision domains, each
with its own capacity.
Switch Creates Four Collision Domains and Four Ethernet
Segments

Switching Logic
 Unicast frames have a unicast address as a destination, these addresses
represent a single device
 broadcast frame has a destination MAC address of FFFF.FFFF.FFFF, this
frame should be delivered to all devices on the LAN
 LAN switches receive Ethernet frames and then make a switching decision:
either forward the frame out some other port(s) or ignore the frame

Switching Logic
 To accomplish this primary mission, transparent bridges perform three
actions:
1. Deciding when to forward a frame or when to filter (not forward) a frame,
based on the destination MAC address
2. Learning MAC addresses by examining the source MAC address of each
frame received by the switch
3. Creating a (Layer 2) loop-free environment with other bridges by using
Spanning Tree Protocol (STP)

The Forward-Versus-Filter Decision
 To decide whether to forward a frame, a switch uses a dynamically built table
that lists MAC addresses and outgoing interfaces
 Switches compare the frame’s destination MAC address to this table to decide
whether the switch should forward a frame or simply ignore it
 If the destination address is a known unicast address , and the outgoing
interface is the same as the interface in which the frame was received, the
switch filters the frame, meaning that the switch simply ignores the frame and
does not forward it
 A switch’s MAC address table is also called the switching table, or bridging
table, or even the Content Addressable Memory (CAM) table, in reference
to the type of physical memory used to store the table

Sample Switch Forwarding and Filtering
Decision

 A switch’s MAC address table lists the location of each MAC relative to that
one switch
 In LANs with multiple switches, each switch makes an independent
forwarding decision based on its own MAC address table. Together, they
forward the frame so that it eventually arrives at the destination
 The forwarding choice by a switch was formerly called a forward-versus-filter
decision, because the switch also chooses to not forward (to filter) frames, not
sending the frame out some ports.

Forwarding Decision with Two Switches

How Switches Learn MAC Addresses
 Switches build the address table by listening to incoming frames and
examining the source MAC address in the frame
 If a frame enters the switch and the source MAC address is not in the MAC
address table, the switch creates an entry in the table
 That table entry lists the interface from which the frame arrived

Switch Learning: Empty Table and Adding Two
Entries

 Switches keep a timer for each entry in the MAC address table, called an
inactivity timer
 The switch sets the timer to 0 for new entries. Each time the switch receives
another frame with that same source MAC address, the timer is reset to 0.
 The timer counts upward, so the switch can tell which entries have gone the
longest time since receiving a frame from that device.
 The switch then removes entries from the table when they become old. Or, if
the switch ever runs out of space for entries in the MAC address table, the
switch can then remove table entries with the oldest (largest) inactivity timers
 Aging time for all MAC addresses can be configured. The default is 300
seconds

Flooding Frames
 Unknown unicast frames: frames whose destination MAC addresses are not
yet in the address table
 Switches flood unknown unicast frames
 Flooding means that the switch forwards copies of the frame out all ports,
except the port on which the frame was received
 If the unknown device receives the frame and sends a reply, the reply frame’s
source MAC address will allow the switch to build a correct MAC table entry
for that device
 Switches also forward LAN broadcast frames, because this process helps
deliver a copy of the frame to all devices in the LAN

Avoiding Loops Using Spanning Tree Protocol
 Without STP, any flooded frames would loop for an indefinite period of time
in Ethernet networks with physically redundant links
 To prevent looping frames, STP blocks some ports from forwarding frames
so that only one active path exists between any pair of LAN segments
 The result of STP is good: Frames do not loop infinitely, which makes the
LAN usable
 However, STP has negative features as well, including the fact that it takes
some work to balance traffic across the redundant alternate links

Network with Redundant Links but Without STP: The Frame Loops
Forever

 To avoid Layer 2 loops, all switches need to use STP
 STP causes each interface on a switch to settle into either a blocking state or
a forwarding state.
 Blocking means that the interface cannot forward or receive data frames,
while forwarding means that the interface can send and receive data frames.
 If a correct subset of the interfaces is blocked, only a single currently active
logical path exists between each pair of LANs
 STP behaves identically for a transparent bridge and a switch

Internal Processing on Cisco Switches
 As soon as a Cisco switch decides to forward a frame, the switch can use a
couple of different types of internal processing variations
 Three types of these internal processing methods are supported in at least
one type of Cisco switch:
1. Store-and-forward
2. Cut-through
3. Fragment-free
 With store-and-forward, the switch must receive the entire frame before
forwarding the first bit of the frame.

 Because the destination MAC address occurs very early in the Ethernet
header, a switch can make a forwarding decision long before the switch has
received all the bits in the frame.
 The cut-through and fragment-free processing methods allow the switch to
start forwarding the frame before the entire frame has been received,
reducing time required to send the frame (the latency, or delay)
 With cut-through processing, the switch starts sending the frame out the
output port as soon as possible. Although this might reduce latency, it also
propagates errors. Because the Frame Check Sequence (FCS) is in the
Ethernet trailer, the switch cannot determine whether the frame had any
errors before starting to forward the frame. So, the switch reduces the frame’s
latency, but with the price of having forwarded some frames that contain
errors.

 Fragment-free processing works similarly to cut-through, but it tries to reduce
the number of errored frames that it forwards.
 One interesting fact about Ethernet CSMA/CD logic is that collisions should be
detected within the first 64 bytes of a frame
 Fragment-free processing works like cut-through logic, but it waits to receive
the first 64 bytes before forwarding a frame.
 The frames experience less latency than with store-and-forward logic and
slightly more latency than with cut-through, but frames that have errors as a
result of collisions are not forwarded
 today’s switches typically use store-and-forward processing

Switch Internal Processing

LAN Switching Features
 Switch ports connected to a single device, providing dedicated bandwidth to
that single device
 Switches allow multiple simultaneous conversations between devices on
different ports
 Switch ports connected to a single device support full-duplex, in effect
doubling the amount of bandwidth available to the device
 Switches support rate adaptation, which means that devices that use different
Ethernet speeds can communicate through the switch (hubs cannot)

Design Choices in Ethernet LANs
Collision Domains
 The different parts of an Ethernet LAN can behave differently, in terms of
function and performance
 The term collision domain referred to an Ethernet concept of all ports whose
transmitted frames would cause a collision with frames sent by other devices
in the collision domain
Collision Domains

Collision Domains
 Only the hub allows a CD to spread from one side of the device to the other
 If PC3 and the LAN switch both enabled half-duplex, which uses CSMA/CD,
they would consider their frames to collide if they were sent and received at
the same time
 A collision domain is a set of network interface cards (NIC) for which a
frame sent by one NIC could result in a collision with a frame sent by any other
NIC in the same collision domain

Broadcast Domains
 Only routers separate the LAN into multiple broadcast domains.
 LAN switches flood Ethernet broadcast frames, extending the scope of the
broadcast domain.
 Routers do not forward Ethernet broadcast frames, either ignoring the
frames, or processing and then discarding some broadcast from some
overhead protocols used by routers.
 bridges act like switches with broadcasts, and hubs repeat the signal, again
not stopping the broadcasts

Broadcast Domains
 Broadcasts sent by a device in one broadcast domain are not forwarded to
devices in
another broadcast domain
 A broadcast domain is a set of NICs for which a broadcast frame sent by one
NIC is received by all other NICs in the same broadcast domain
Broadcast Domains

The Impact of Collision and Broadcast Domains on LAN
Design
 For a single collision domain:
 The devices share the available bandwidth
 The devices might inefficiently use that bandwidth because of the effects
of collisions, particularly under higher utilization
 When a host receives a broadcast, the host must process the received frame.
 This means that the NIC must interrupt the computer’s CPU, and the CPU
must spend time thinking about the received broadcast frame
 Broadcasts do require all the hosts to spend time processing each
broadcast frame
 Using smaller broadcast domains can also improve security, because of
limiting broadcasts and because of robust security features in routers

The Impact of Collision and Broadcast Domains on LAN
Design
Benefits of Segmenting Ethernet Devices Using Hubs, Switches,
and Routers

Virtual LANs (VLAN)
 A LAN consists of all devices in the same broadcast domain.
 With VLANs, a switch groups interfaces into different VLANs (broadcast
domains) based on configuration, with each interface in a different VLAN
 Essentially, the switch creates multiple broadcast domains by putting some
interfaces into one VLAN and other interfaces into other VLANs
 So, instead of all ports on a switch forming a single broadcast domain, the
switch separates them into many, based on configuration
 Without VLANs, a switch considers all interfaces on the switch, and the
devices connected to those links, to be in the same broadcast domain

Virtual LANs (VLAN)
Sample Network with Two Broadcast Domains and No
VLANs
Sample Network with Two VLANs Using One
Switch

Campus Design Terminology
 The term campus LAN refers to the LAN created to support larger buildings, or
multiple buildings in somewhat close proximity to one another
 Cisco uses three terms to describe the role of each switch in a campus
design:
1. Access
2. Distribution
3. Core
 The roles differ based on whether:
 The switch forwards traffic from user devices and the rest of the LAN
(access)
 The switch forwards traffic between other LAN switches (distribution and
core)
 Using designs that connect a larger number of access switches to a small

Campus LAN with Design Terminology Listed

Access switches:
 Connect directly to end users, providing user device access to the LAN.
 Send traffic to and from the end-user devices to which they are connected and
sit at the edge of the LAN
Distribution switches:
 Provide a path through which the access switches can forward traffic to each
other.
 Each of the access switches connects to at least one distribution switch,
relying on distribution switches to forward traffic to other parts of the LAN
 Most designs use at least two uplinks to two different distribution switches for
redundancy
Core switches: The largest campus LANs often use core switches to forward
traffic between distribution switches

Ethernet LAN Media and Cable Lengths
 When designing a campus LAN, an engineer must consider the length of each
cable run and then find the best type of Ethernet and cabling type
 10BASE-T, 100BASE-T, and 1000BASE-T have the same 100-meter cable
restriction, but they use slightly different cables
 The EIA/TIA defines Ethernet cabling standards, including the cable’s quality
 Each Ethernet standard that uses UTP cabling lists a cabling quality category
as the minimum category that the standard supports:
 10BASE-T allows for Category 3 (CAT3) cabling or better
 100BASE-T calls for higher-quality CAT5 cabling
 1000BASE-T requires even higher-quality CAT5e or CAT6

 Optical cables support a variety of much longer distances than the 100
meters supported by Ethernet on UTP cables
 Optical cables experience much less interference from outside sources as
compared to copper cables
 The type of optical cabling can also impact the maximum distances per cable:
 Multimode fiber supports shorter distances, but it is generally cheaper
cabling and it works fine with less-expensive LEDs.
 Single-mode fiber supports the longest distances but is more
expensive. Often use laser-based hardware

Ethernet Types, Media, and Segment Lengths
(Per IEEE)

Autonegotiation
 Ethernet devices on the ends of a link must use the same standard or they
cannot correctly send data
 IEEE autonegotiation (IEEE standard 802.3u) defines a protocol that lets the
two UTP-based Ethernet nodes on a link negotiate so that they each choose
to use the same speed and duplex settings.
 The protocol messages flow outside the normal Ethernet electrical frequencies
as out-of-band signals over the UTP cable
 Each node states what it can do, and then each node picks the best
options that both nodes support:
 The fastest speed and the best duplex setting, with full-duplex being better
than half-duplex

Autonegotiation
Many networks use autonegotiation every day, particularly between user devices
and the access layer LAN switches
IEEE Autonegotiation Results with Both Nodes Working
Correctly

Autonegotiation Results When Only One Node Uses
Autonegotiation
 Most Ethernet devices can disable autonegotiation, so it is just as important
to know what happens when a node tries to use autonegotiation but the node
gets no response
 If autonegotiation enabled on both ends of the link, the nodes should pick the
best speed and duplex. However, when enabled on only one end, many issues
can arise: The link might not work at all, or it might just work poorly
 IEEE autonegotiation defines some rules that nodes should use when
autonegotiation fails:
 Speed: Use your slowest supported speed (often 10 Mbps)
 Duplex: If your speed = 10 or 100, use half-duplex; otherwise, use full-
duplex
 Cisco switches can actually sense the speed used by other node, even

Autonegotiation
IEEE Autonegotiation Results with Autonegotiation Disabled on
One Side

Autonegotiation
 PC1 shows a classic and unfortunately common end result: a duplex
mismatch
 The two nodes can send data However, PC1, using full-duplex, does not
attempt to use CSMA/CD logic and sends frames at any time.
 Switch port F0/1, with halfduplex, does use CSMA/CD. As a result, switch port
F0/1 will believe collisions occur on the link, even if none physically occur
 The switch port will stop transmitting, back off, resend frames, and so on. As
a result, the link is up, but it performs poorly
 when both devices are attempting to transmit at the same time, the packet
sent by the full-duplex end will be discarded and lost due to an assumed
collision and the packet sent by the half duplex device will be delayed or lost

Autonegotiation and LAN Hubs
 Hubs do not react to autonegotiation messages, and they do not forward the
messages.
 As a result, devices connected to a hub must use the IEEE rules for choosing
default settings, which often results in the devices using 10 Mbps and
halfduplex
IEEE Autonegotiation with a LAN Hub

Building Ethernet LANs with Switches
References
1) Cisco Systems, Inc, www.cisco.com/
2) Wendell Odom ,”Cisco CCENT/CCNA ICND1 100-101 Official Cert Guide”,
Cisco Press, USA, 2013

CCNA R&S-07-Building Ethernet LANs with Switches

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (11)

More from Amir Jafari

More from Amir Jafari (7)

Recently uploaded

Recently uploaded (20)

CCNA R&S-07-Building Ethernet LANs with Switches

Editor's Notes