CompTIA Network+

EXAM
N10-008
N10-008 Exam Domains
1.0 Networking Fundamentals 24%
2.0 Network Implementation 19%
3.0 Network Operations 16%
4.0 Network Security 19%
5.0 Network Troubleshooting 22%
Total 100%
Networking Fundamentals

DOMAIN 1.0
MODULE 1
Networking Fundamentals Topics
Intro to Networking
Common Terminology
OSI Model
DOD Model
TCP/IP
Network Topologies
Network Types and Characteristics
Intro to Networking
What is a Network?
Two or more computers connected together
The computers can be any type of computing device
The connection can be wired or wireless
Why Have a Network?
Share data
Remote communication
Share resources such as printers, faxes, databases, and services
Distribute a computing workload
◦ Sensor – Monitor
◦ Client – Server
◦ Multiple facilities working together

Cost effectiveness and reliability

How Computers Communicate on a
Network
Both Sides Need:
Applications that want to talk to each other
How Computers Communicate on a
Network
Both Sides Need:
Applications that want to talk to each other

Common protocol (language)

How Computers Communicate on a
Network
Both Sides Need:
Applications that want to talk to each other

Common protocol (language)

Network Interface to connect to network

How Computers Communicate on a
Network
Both Sides Need:
Applications that want to talk to each other

Common protocol (language)

Network Interface to connect to network

Transmission Media (wired or wireless)

How Computers Communicate on a
Network Two hosts want to communicate

Host 1 Host 2
How Computers Communicate on a
Network Usually, both hosts have applications
that want to communicate

App App

Host 1 Host 2
How Computers Communicate on a
Network Usually, both hosts have applications
that want to communicate

App But most apps are not designed to use App

a network directly

Host 1 Host 2
How Computers Communicate on a
Network Usually, both hosts have applications
that want to communicate

App But most apps are not designed to use App

a network directly

Host 1 They need the operating system with Host 2

its networking services to help them
Basic Network Components
Networking has the following components:

App Applications that want to App

communicate

Host 1 Host 2
Basic Network Components
Networking has the following components:

App Applications that want to App

communicate
Protocol A common language (protocol) Protocol
Host 1 Host 2
Basic Network Components
Networking has the following components:

App Applications that want to App

communicate
Protocol A common language (protocol) Protocol
Host 1 Host 2

NIC A network interface NIC

Basic Network Components
Networking has the following components:

App Applications that want to App

communicate
Protocol A common language (protocol) Protocol
Host 1 Host 2

NIC A network interface NIC

Transmission media to physically connect them

Basic Network Components
Let’s look again with examples:

Web Browser Applications that want to communicate Web Server

Host 1 Host 2
Basic Network Components
Let’s look again with examples:

Web Browser Applications that want to Web Server

communicate
TCP/IP A common language (protocol) TCP/IP
Host 1 Host 2
Basic Network Components
Let’s look again with examples:

Web Browser Applications that want to communicate Web Server

TCP/IP A common language (protocol) TCP/IP

Host 1 Host 2

Intel Pro
1000 MB
A network interface Intel Pro
1000 MB
Basic Network Components
Let’s look again with examples:

Web Browser Applications that want to communicate Web Server

TCP/IP A common language (protocol) TCP/IP

Host 1 Host 2

Intel Pro
1000 MB
A network interface Intel Pro
1000 MB

Transmission media to physically connect them

CAT 6 cable
Client / Server Usually a
dedicated
computer acts
as the server

Web Browser Web Server

(Client) (Server)

Client initiates the connection

Host 1 Host 2
Client / Server Usually a
dedicated
computer acts
as the server

Web Browser Web Server

(Client) (Server)

Client initiates the connection

Host 1 Host 2
Server waits for clients to connect
• Can accept or reject the connection
attempt
Client / Server Usually a
dedicated
computer acts
as the server

Web Browser Web Server

(Client) (Server)

Client initiates the connection

Host 1 Host 2
Server waits for clients to connect
• Can accept or reject the connection
attempt
“Client” and “Server” can refer to:
• The device itself
• A process on the device
Peer-to-Peer

App App

No dedicated server
Host 1 Host 2
Both sides act as client and server
Common
Terminology
What is an IP Address?
A number that identifies a node on the network
Can be changed
◦ User configures a different address in network settings
◦ Device automatically obtains a new address when connected to a different network
A device can have more than one IP address
◦ One per interface
◦ Alternate IP addresses on the same interface
Each IP address must be unique on the network so there is no conflict
Analogous to a phone number
Examples:
IPv4 - 192.168.1.10
IPv6 - 2601:140:8780:43f0:2d1e:9ebb:92f0:d6e9
What is a MAC Address?
The physical address of the network interface
◦ Also known as the “Burned In Address” (BIA)
◦ Does not change

Assigned by the NIC vendor

One per interface
Analogous to a cell phone MEID or IMEI number, or a device serial number
Examples: BC-85-56-F3-13-02, BC:85:56:F3:13:02, BC85.56F3.1302
What are Source and Destination?
Source is the node that is transmitting/sending
Destination is the intended recipient of the transmission
A Client and Server take turns transmitting to each other
◦ Thus they alternate between being source and destination
 What is a Unicast, Multicast, Broadcast?
A node transmits a packet on the network
Depending on the destination address, one or more nodes will receive the packet

Unicast Multicast Broadcast

One to one One to Some One to All
What is a Protocol?
Set of rules or “language” for communication
Can exist at any level/layer of networking
A host will use several protocols to make a connection on the network
What is a Port?
A number that represents an application (process) on the network
Assigned by the operating system
Every process that accesses the network has its own unique port number on that device
What is a Socket?
A socket is a port that is in use
It is a combination of protocol, IP address, and port
This combination uniquely identifies the connection on the network
What is a Socket?
A socket is a port that is in use
It is a combination of protocol, IP address, and port
This combination uniquely identifies the connection on the network

Example:

TCP 192.168.1.5:80

Protocol IP Address Port

OSI Model
OSI Model
7 Application
OSI Model
7 Application

6 Presentation
OSI Model
7 Application

6 Presentation

5 Session
OSI Model
7 Application

6 Presentation

5 Session

4 Transport
OSI Model
7 Application

6 Presentation

5 Session

4 Transport

3 Network
OSI Model
7 Application

6 Presentation

5 Session

4 Transport

3 Network

2 Data Link
OSI Model
7 Application

6 Presentation

5 Session

4 Transport

3 Network

2 Data Link

1 Physical
Protocol Data Unit (PDU)
A specific block of information transferred over a network
This term is used in reference to the OSI model, describing the data at each layer:
◦ Application = data
◦ Presentation = data
◦ Session = data
◦ Transport = segment (TCP), datagram (UDP)
◦ Network = packet
◦ Data Link = frame
◦ Physical = bit
Layer 7 – Application Layer

“Customer service counter” that applications use to request network services

User can typically interact at this layer
Applications connect by speaking a language (protocol)
Common Layer 7 protocols include:
◦ SMTP, POP3, IMAP4, HTTP, HTTPS, RDP, DNS, DHCP, SMB, NFS, FTP, TFTP, Telnet, SSH, SIP, NTP,
SNMP, LDAP

Firewalls, proxies At this layer,

data is called
“data”
Layer 6 – Presentation Layer
Both sides agree on a common data format At this layer,
User may or may not be able to interact with this layer
data is stilled
◦ Might be prompted to install a browser plugin to watch a video called “data”
Common formats include:
◦ Multimedia formats – JPG, PNG, GIF, MP3, MP4, MKV, MOV, WAV, PDF…
◦ Encryption algorithm and bit size – DES, AES, MD5, SHA-1, 160 bit, 128 bit…
◦ Compression – H.264, H.263, MPEG-4, MPEG-2, AAC,
◦ Character sets – ASCII, Unicode, EBCDIC

Firewalls
Layer 5 – Session Layer
Keeps separate conversations separate
It is here that a host has the first concept of communicating with another host
Usually done by assigning ports (source and destination) to a conversation
Can also be NetBIOS named pipes or Unix sockets At this layer,
data is stilled
Firewalls, packet filtering routers, multi-layer switches
called “data”
SOCKS proxies
Layer 4 – Transport Layer
Pivotal layer – abstracts the mechanics of the network from the higher layers
Starts, manages, and tears down the session
First layer to encapsulate the data payload with a header
TCP, UDP
Firewalls, packet filtering routers, multi-layer switches
Layer 4 – Transport Layer (cont’d)
TCP:
◦ Breaks up data into manageable pieces for transmission
◦ Adds sequence numbers to each segment for reassembly at other end Data is called a
◦ Embeds the source and destination ports into its header “Segment”
◦ Establishes the session with a handshake
◦ Provides error correction and flow control during session
◦ Tears down session with a handshake Data is called
UDP: a “Datagram”
◦ Embeds the source and destination ports into its header
◦ Depends on the application for session establishment, management, payload length, error correction,
flow control, and tear down
Layer 3 – Network Layer
Encapsulates Layer 4 into a packet
Adds a logical address (usually IP address)
Data at this
Chooses the best route
layer is called
IP, ICMP, IGMP a “packet”
Routers, Firewalls, multi-layer switches
Layer 2 – Data Link Layer
Data at this
Adds a physical source and destination address
layer is called
Encapsulates Layer 3 packet into a frame a “frame”
◦ Adds both a header and trailer
Formats the frame to be suitable for transmission media
Checks incoming frames for errors
◦ Discards frames that do not pass a simple cyclical redundancy check (CRC)
◦ Depends on Layer 4 to retransmit discarded data
Has two sub layers:
◦ Logical Link Control (LLC) – describes the Layer 3 payload
◦ Media Access Control (MAC) – adds the physical addresses
ARP, Ethernet, Token Ring, PPP, HDLC, Frame Relay
Switches, bridges
Layer 1 – Physical Layer
Data at this layer
is called “bits”
Actual transmission of the frame as 1’s and 0’s
All electrical and mechanical aspects of the transmission
Includes connectors, wiring types, wireless technologies, baseband, broadband, modulation,
speed, bandwidth, clock rate, voltages, frequencies, power levels…
Hubs, repeaters, patch panels, network interface cards, RJ-45, STP, UTP, thicknet/thinnet coax,
CAT 3/5/5e/6/6a/7, fiber optic, 2.4/5 GHz, Wi-Fi channels, ZigBee, Z-Wave, infrared and other
wireless technologies
Data Encapsulation / Decapsulation
Encapsulation
Layers 4, 3, 2 on the sender add a header
◦ Description of the data and/or protocol
◦ Layer 2 also adds a trailer

Decapsulation:
Corresponding layers on the receiver remove the header
◦ Refer to the information in the header/trailer
◦ Pass the PDU up to the next layer
 OSI Encapsulation/Decapsulation
Sender Receiver

7 Application Each layer is a payload Application 7

of the layer below it
6 Presentation Presentation 6

5 Session Data Data Session 5

Transport Transport
4 Encapsulation Data Data 4
TCP, UDP

Decapsulation
TCP, UDP
Network Network
3 Data Data 3
IP IP
Data Link Data Link
2 Data Data 2
Ethernet Ethernet

1 Physical 00001011 10101011 11100110 01011010 11001001 Physical 1

 OSI Model Summary
7 Application Customer service counter for app HTTP, HTTPS, FTP, TFTP, SMTP, POP3, IMAP4, SMB,
to request network services NFS, RDP, LDAP, DNS, DHCP, SSH, Telnet, SNMP …
Both sides agree on common All encryption, multimedia formats, character sets,
6 Presentation
data format encryption
Keep separate conversations Ports, named pipes, NetBIOS, RPC
5 Session
separate
Establish, manage, tear down a TCP, UDP
4 Transport
connection
Add logical address, choose the IP, ICMP, IGMP
3 Network best route
Format the data for transmission, LAN and WAN protocols, ARP
2 Data Link add physical address
Actually transmit packet as 1’s All physical and electrical characteristics of
1 Physical and 0’s connectors and transmission media
DoD Model
What is the DoD Model?
Developed by the US Department of Defense
4 layer model
Its layers map to the OSI model
Sometimes referred to as the TCP/IP Model
◦ Specifically uses the TCP/IP suite
OSI – DoD Layer Mapping
7 Application

6 Presentation Application

5 Session

4 Transport Host-to-Host

3 Network Internet

2 Data Link
Network Access
1 Physical
TCP/IP Suite
Application • Data Payload
Transmission Control Protocol /
Internet Protocol
• Session Establishment /
Protocols of the DoD Model Host-to-Host Management
TCP, UDP • Sequencing
6 core protocols:
◦ TCP, UDP, IP, ICMP, IGMP, ARP

All but ARP have an IANA assigned Internet • Logical Addressing

protocol number (protocol ID) IP, ICMP, IGMP • Routing

Many auxiliary protocols also exist

at the various layers • Physical Addressing
Network Access
ARP • Framing/CRC/MTU
• Actual transmission
What is a Connection-Oriented Protocol?
Attempts to ensure reliability and completeness of transmission
Uses a handshake to create and end a session
◦ Like sending registered mail
Keeps track of the conversation
Ensures that the other side is responding
◦ Is the other side acknowledging received packets?
◦ Resends packets that are not acknowledged
◦ Acknowledgements and resends add overhead and slow the conversation
Responds to requests from the other side
◦ Receiver informs sender how many packets it can receive at a time
◦ Sender speeds up or slows down the transmission rate accordingly
Used when reliability is more important than performance
What is a Connectionless Protocol?
Makes no attempt to ensure completeness of the transmission
No handshake
◦ Does not even know or care if the recipient is online or offline
◦ Like sending post cards

Expects higher level protocols or the app to request resends if some of the data did not arrive
◦ Assumes lost packets are not important or will be re-transmitted

Used when performance is more important than reliability

 Connection-Oriented vs Connectionless
Connection-Oriented Connectionless
TCP UDP, IP, ICMP, IGMP, ARP
Handshake to set up / tear down session No handshake
Flow control / error correction No flow control / no error correction
“Reliable” “Unreliable” or “best effort”
Focus on receiving all the data Focus on speed/performance
Used for transferring files: • Used for communications that are
• web pages time sensitive, and/or can tolerate
• emails some loss:
• file transfers • Realtime voice and video
• video-on-demand • SNMP, DNS, DHCP
• remote control
TCP, UDP, IP
Transmission Control Protocol (TCP)
Host-to-host Layer (Layer 4) protocol
Payload of IP
Embeds source and destination ports in its header
Provides reliable, connection-oriented communication over IP networks between two endpoints
Connection oriented
◦ Attempts to guarantee delivery
Transmission Control Protocol (TCP)(cont’d)

Data is broken into smaller segments

◦ Identified by sequence #

Uses a receive window (sliding window)

◦ Tells the sender how big the receiver’s buffer is from segment to segment

Session established by three-way handshake (SYN – SYN/ACK – ACK)

Session closed by a four-way handshake (FIN-ACK-FIN-ACK)
Protocol ID 6
TCP Header
TCP Flags
Indicate the purpose of the TCP segment
Every TCP segment will have at least one flag raised

Flag Description
URG Tells receiver to prioritize this data
ACK Acknowledge
All TCP segments (except very first and last will have an ACK)
PSH Tells receiver to directly send this data to the application
RST Tells receiver that the sender has abruptly ended the conversation
SYN Used in 3-way handshake to start conversation
FIN Used in 4-way handshake to end conversation
User Datagram Protocol (UDP)
TCP’s “Little Brother”
Host-to-Host Layer (Layer 4) protocol
Connectionless (“unreliable” or “best effort”)
A payload of IP
Embeds source and destination ports in its header
No session establishment
User Datagram Protocol (UDP) (cont’d)
This protocol sends short packets of data, called datagrams
Ideal for network applications where latency is critical but loss is not:
◦ Gaming, real-time voice and video
◦ SNMP, DNS queries and DHCP
◦ Some applications (especially voice) use various techniques to make up for any loss

Protocol ID 17
UDP Header
Internet Protocol (IP)
Internet Layer (Layer 3) connectionless protocol
Can be a payload of nearly any link-layer protocol
Protocol ID 4
IPv4:
◦ Still most widely used version
◦ Uses 32-bit logical addressing to identify source and destination host
◦ Address examples: 10.0.0.207, 172.16.54.3, 207.45.168.4, 198.32.8.75
IPv6:
◦ Allows longer addressing and vast improvements over IPv4
◦ Steadily gaining in popularity and use
◦ Address examples: 2601:140:8780:43f0::4762, fe80::10:18ff:fe12:2001
IPv4 Header
IPv6 Header
320 bits (40 Bytes)
IPv6 Header Extensions

IPv6 Header Extension Extension Extension

Upper Layer
Header 1 Header 2 Header n
Data
(TCP, UDP)
Next Header Next Header Next Header Next Header
ICMP, IGMP, ARP
Internet Control Message Protocol (ICMP)
An error-reporting protocol used by network devices (e.g. routers) to generate error messages
and manage traffic flow
Layer 3 payload of IP
Any IP network device can send and receive ICMP messages
Used by PING application
Protocol ID 1
ICMP Message Codes
ICMP uses different codes to identify the purpose of the message:
◦ Echo request
◦ Echo reply
◦ Destination unreachable
◦ Source quench
◦ Redirect
◦ Router solicitation
◦ Router advertisement
◦ Time exceeded
ICMP Message Format
Applications that Use ICMP
PING
◦ An application that uses ICMP to prove Layer 3 connectivity
◦ NOT a protocol – do not confuse with ICMP

Traceroute
◦ Stream of ICMP echo requests OR UDP datagrams with limited TTL
◦ Router that discards the expired packet and sends Expired in Transit message while identifying itself
◦ Microsoft traceroute application is called tracert
Internet Group Management Protocol
(IGMP)
Used by hosts to notify routers that they are still interested in receiving multicasts from
upstream server
◦ Routers do not forward multicasts by default
◦ Host must let upstream router know that it wants the multicast

Protocol ID 2
IGMP Operation
Router needs to know
if downstream devices
still want the
transmission

Switch listens for IGMP

requests from hosts –
only forwards IGMP to
hosts that request it

Hosts that want to

receive the multicast
send IGMP messages
up to the router
IGMP Packet Format
Address Resolution Protocol (ARP)
Network Access Layer (Layer 2) connectionless protocol
Used to map MAC addresses to IP addresses
Sends Layer 2 broadcast (FFFFFFFFFFFF) querying all listening nodes to identify which one is
using the specified IP address
Mappings are temporarily stored in the device’s ARP cache
Used in Ethernet and Wi-Fi networks
Does not have an assigned protocol number
ARP Message Format
Network Topologies
Network Topology
Describes the layout of a network
Determines how devices connect and communicate
Topologies are either physical or logical
◦ Physical – actual physical layout/physical connections between devices on a network
◦ Logical – how data moves from one device to another

Physical and logical topologies need not be the same

Point-to-Point
Only two nodes on a link
Examples:
◦ WAN link between two locations
◦ Line-of-sight wireless between two buildings
◦ Uplink/trunk link between two switches
◦ Two PCs connected by a crossover cable
Point-to-Point Topology Examples
Bus
Comprised of a single cable that all devices tap into
Cable ends have to be specially terminated so signal does not reflect back onto bus
Original Ethernet networks used a bus topology
If the main cable breaks the network goes down
◦ It can be difficult to troubleshoot
Star
Network devices are connected to a central device/node
called a hub
Nodes communicate across the network by passing data
through the hub
Also known as:
◦ Hub-and-Spoke
◦ Point-to-multipoint (WANs or wireless)

Advantage – if one device or link malfunctions, the

remainder of network still functions
Disadvantage - if the hub fails, the whole network goes down
Star Topology Examples
Point-to-Multipoint Wireless Example
Ring
Network devices are arranged in a loop
Each node acts as a repeater
Advantage:
◦ Regenerates signal when passing data through each device
◦ Thus can support a larger network

Disadvantage:
◦ Can be slower than star
◦ If one device goes down all of devices will be impacted
Dual Ring
Used in FDDI MANs
Redundant rings provide fault tolerance
If the rings are broken at one point, nodes on either side use both rings to close the loop
Mesh
Nodes are connected to more than one node
◦ Redundancy
◦ Fault tolerance

Can be thought of as a redundant star

Two types of mesh topologies:
◦ Full mesh - Every node connects to every other node
◦ Full redundancy
◦ Expensive, but good for backbone (core) network
◦ Partial mesh – Some nodes have links to more than one node
◦ Is less expensive with less redundancy
◦ Good for devices that connect to backbone network
Hybrid
Any combination of the various topologies to create a larger network

Point-
to-Point
STAR STAR
link
Network Types
Local Area Network (LAN)
Limited to a small geographical region
◦ A floor, building, or small campus

Uses LAN-based or dual-use network protocols/technologies

◦ Ethernet
◦ Wireless
◦ Token Ring
◦ ATM

One organization usually owns all of the equipment/infrastructure

Metropolitan Area Network (MAN)
A network that connects users with computer
resources in a geographic area or region
◦ Typically around a town/municipality (5 - 50 km)
◦ Larger than a LAN
◦ Smaller than a WAN

Uses MAN-specific protocols

LAN (high) speed on fiber optic cable
Often used by companies to connect multiple
sites around town
Typically owned by a consortium of users
◦ Or a single provider that sells the service to users
MAN Protocols
Asynchronous Transfer Mode (ATM)
Fiber Distribution Data Interface (FDDI)
Switched Multi-megabit Data Service (SMDS)
Multiprotocol Label Switching (MPLS)
Metropolitan Ethernet (Metro Ether)
Wide Area Network (WAN)
Connects remote locations
◦ Across towns, states, even continents
◦ Owned by one or more service providers

Likely to use different network protocols in various network segments

Customers pay for the provider to connect their remote offices
Traditionally much slower than LANs or MANs
The Internet is the largest example of a WAN
The Internet is the Ultimate WAN
Comprised of thousands of telecom networks
all connected together
When a telecom provides Internet service, it is
called an “ISP” (Internet Service Provider)
ISPs are identified by their Autonomous System
(AS) Number
Permits billions of devices and people to
communicate across the globe
Wireless LAN (WLAN)
Uses Wi-Fi instead of Ethernet to connect devices
Can be connected to an Ethernet LAN
Personal Area Network (PAN)
A computer network organized around an individual person
Set up for personal use only
Can include a computer, phone, printer, tablet, wearables, and other personal devices
Can use many technologies (wired or wireless)
◦ Traditionally was Bluetooth-based (802.15)
◦ Can also include other wireless technologies

Also known as a piconet

Campus Area Network (CAN)
A large LAN, covering a campus of buildings
Likely to have a high-speed fiber optic backbone
Smaller MAN networks are sometimes referred to as “CANs”
Storage Area Network (SAN)
Most commonly used storage architecture in business
Specialized network segment connects servers to centralized, dedicated storage devices
Provides high performance and low latency
Can use:
◦ Specialized fibre channel cable and switches
◦ Existing Ethernet network

Can span multiple sites

Can play an important role in Business Continuity
◦ Important data is all in one place
◦ Easier to secure, back up, recover, and manage
Typical SAN Architecture
Network
Characteristics
What are Baseband and Broadband?
Baseband
◦ Data signal is placed directly on transmission medium
◦ Generally, transmission medium can only carry one signal at a time
◦ Distance limited by quality of transmission medium and original signal strength
◦ Most LAN transmissions are baseband
◦ Analogous to sound carried on a microphone cable or speaker wire

Broadband
◦ Digital data signal is “piggy-backed” (modulated) on top of a more powerful analog carrier signal
◦ Multiple carrier frequencies allow multiple signals to share the same transmission medium
◦ Distance is limited by strength of the carrier signal and type of transmission medium
◦ DSL and cable modem
Multiprotocol Label Switching (MPLS)
Internal network used by service providers to:
◦ Create private WAN links for customers
◦ Connect data centers to branch offices

Not a service, but a routing technique

◦ Directs traffic from one node to the next based on short path labels instead of long network addresses
◦ Used to speed up the flow of traffic in a network and to ensure reliable packet delivery of data packets

Operates between traditional OSI Layer 2 and 3

◦ Sometimes referred to as Layer 2.5 protocol

A trusted go-to connectivity option for organizations for decades

◦ Still in use today
MPLS Process
1. The packet enters an MPLS network (domain)
2. The first router (ingress router) performs a traditional route table lookup
3. Instead of finding a next-hop, it finds the final destination router
4. Then it finds a pre-determined path from “here” to that final router
5. Based on this information, the router applies a “label” (or “shim”) to the packet
6. The other routers along the path use the label to forward the packet to the next
router in the path
◦ No need to perform any additional IP lookups
◦ They know what the path should be, and keep the packet on that path
7. The final router removes the label
8. The packet is delivered to the destination via normal IP routing
MPLS Header
MPLS Domain Example
MPLS Advantages
Much faster than traditional routing
Reliable and secure
Built-in traffic engineering
Can have pre-calculated backup routes handy
Offers customers a private connection with end-to-end QoS
Works with IP, ATM, Frame Relay, Ethernet or SONET
MPLS providers also can deliver more comprehensive service-level agreements than providers
that offer connectivity over public links
MPLS Disadvantages
Solutions are geographically limited to locations where dedicated MPLS circuits are available
Complex
Time-consuming to manage
Somewhat inflexible
Costly to upgrade and scale
High bandwidth usage compared to the newer SD-WAN
Software Defined WAN (SD-WAN)
SD-WAN is a software-defined approach to managing the WAN
The latest incarnation of WAN technology, primarily for enterprises
Evolved out of MPLS
Developed to address some of the limitations of traditional WANs
More economical and flexible than MPLS
Traditional WAN Challenges
Traditionally, users connected back to the corporate data center (“back haul”) to access business
applications
◦ If you want to centrally apply and manage organizational policies, you have to route the traffic to the
corporate LAN, even if it is ultimately destined to go back out to the Internet or a cloud service
◦ Results in delayed response times and poor network performance
MPLS not originally designed to handle cloud-based SaaS and IoT traffic
◦ Cloud-based SaaS allows users to connect to the same apps from anywhere, using their local ISP
◦ The problem is, you don’t have centralized control over who connects to what and how
Configuration in traditional WAN architecture is distributed
◦ Housed locally on individual physical routers
◦ Changes and new deployments are time consuming
◦ Configuration usually done manually on a per-device basis by on-site IT engineers
SD-WAN Architecture
Optional master software that coordinates multiple SD-WANs
◦ Automates onboarding new networks and devices into the entire SD-WAN fabric
Orchestrator ◦ Makes it easy for admins and customers to set overarching policy and service requirements
◦ Directs and coordinates the control planes of participating SD-WANs
◦ Typically includes a self-service portal for customers and admins

Centralized software that manages the SDWAN

◦ Takes direction from the orchestrator
◦ Coordinates underlying networks into a single WAN
Control Plane ◦ Allows admins to see and manage all traffic through a single pane-of-glass
◦ Continuously monitors, making real-time routing decisions
◦ Simultaneously deploys configuration and policy to underlying networks

The physical part of the SD-WAN

Forwarding Plane ◦ Comprised of existing WANs from various operators
◦ MPLS, 4G/5G/LTE, broadband Internet, etc.

Where network endpoints reside

Edge ◦ Clients, servers, storage, customer firewall
◦ Can be a provider’s cloud, branch office, company headquarters, datacenter, teleworker’s home
SD-WAN Control and Forwarding Functions
Control Plane (“overlay”) Forwarding Plane (“Underlay”)
◦ Centralized software application ◦ Physical infrastructure including switches,
◦ Creates a virtual network (“overlay”) on top of routers, and links
the physical circuits (“underlay”) ◦ Tunneled traffic is moved across any
◦ Monitors all circuits in the underlay combination of traditional networks
◦ MPLS, cellular, broadband Internet
◦ Controls routing decisions in real-time
◦ Traffic is routed based on policies created by
◦ Continuously chooses the fastest and best-
network administrators
performing available route
◦ Creates tunnels for traffic to move across the
underlay
◦ Tunneling protocols are typically GRE, mGRE,
IPSEC or Secure Vector Routing (SVR)
SD-WAN Example
OSS = Operational Support System
BSS = Business Support System
CE = Customer Edge
SD-WAN Advantages
Dramatic savings over traditional WANS
Constantly, dynamically choosing the best route
Can use cheaper circuits that still maintain the desired performance and security
Increased network visibility for IT administrators
More control over endpoints
Enables connections between geographically remote sites
Reduces hardware dependencies
Eliminates lock-in to service provider solutions
SD-WAN Disadvantages
Without a good orchestrator, can be complex to adopt
Solutions often lack natively integrated security
◦ Can put organizations at risk for cyberattacks
Multipoint Generic Routing Encapsulation (mGRE)
Point-to-multipoint tunneling between routers
◦ Originally, GRE was point-to-point
Allows a single GRE interface to support multiple IPsec tunnels.
Packets are encapsulated (hidden) inside IP packets
Hidden packets can be other Layer 3 payloads
◦ Multicasting
◦ IPv6 Internet
◦ IPv4
Encryption can be added
Used in MPLS and SD-WAN networks
Tunneling = to hide a packet inside a packet so it can be transmitted across
a certain type of network
Service-related Entry Point
Demarcation
◦ Where outside phone/Internet service connects
to inside wiring
◦ At home – outside jack
◦ At business – telco room/wiring closet
(sometimes server room)
◦ Aka “demark”

Smart Jack
◦ Placed between demark and customer wiring
◦ Business installations
◦ Creates a loopback for signal to return to telecom
◦ Used for remote line diagnostics
Virtual Networks
Network (and its attached devices) are not physical
◦ Software processes inside a host computer
◦ Like a little world inside the host
◦ Virtual machines connect to virtual networks inside the host

Connectivity can be restricted to inside the host, or allowed to connect to the outside world
Can also be hosted in a provider’s cloud
◦ Connected to across the Internet or VPN by outside devices
Virtual Network Concepts
vSwitch
◦ Provides switch functionality to virtual machines (VMs)
◦ A host can have many vSwitches, connecting groups of VMs together

Virtual network interface card (vNIC)

◦ The “network interface” that allows a VM to connect to a vSwitch

Hypervisor
◦ A layer of functionality in the host that allows virtual devices to share the host’s physical hardware
Host-based Virtual Network Example

vNIC

vSwitch
 Cloud-based Virtual Network

Teleworker

Office
Provider Links
Link Type Description Typical Max Speed
Satellite • Used where other connection types not available Hundreds of gigabits
• One satellite covers a large area
• Originally high latency, now dramatically reduced using AI
Digital subscriber • Uses existing telephone landline Download 45 mb/s
line (DSL) • Carries traditional Point-to-Point dialup protocol over ATM Upload 7 mb/s
• Requires special modem & user authentication
• PPP over Ethernet at customer site
Cable • Uses existing cable TV network 300 mb/s
• Transmit and receive data carried on two premium TV
channels
Leased line • Dedicated to a single customer 1.544 mb/s
• Primary Rate ISDN, T3, T1, or fractional T1 44.736 mb/s
• PPP or HDLC
Metro-optical Optical fiber connecting to provider’s MAN 100 mb/s