Week 4: Network Layer Functions and Protocol

DHCP

(Dynamic Host Control Protocol) A Protocol used to automatically assign IP addressing information to IP network computers.

IPv4 vs IPv6

- IPv4 has addresses that are 4 bytes long - IPv6 has addresses that are 16 bytes long. The reason we are moving to IPv6 is that we were running out of IPv4 addresses. A 4-byte address has 32 bits, which means there are about 4 billion different addresses. That sounds like a lot, but think about how many different devices you have that connect to the Internet, each of which has an address.

Three Common Routing Protocols (and bonus)

- RIP - # of hops - OSFP and EIGRP - # of hops, traffic report - (bonus) BGP is a special protocol that is used between two different networks; for example, when the core router at Indiana University talks to the router at our Internet Service Provider.

IPv6 Protocol PDU Format

- Version - Priority - Flow Name - Total Length - Next Header - Hop Limit - Source Address - Destination Address - Data

ARP

Address Resolution Protocol. An Internet protocol used to map an IP address to a MAC address. Defined in RFC 826.

Every computer on a multipoint circuit (and many on point-to-point circuits) has a data link layer address (often an Ethernet address)

An Ethernet address is written as six pairs of numbers and letters, separated by dashes (e.g., 90-32-4B-F1-0A-54). The purpose of the data link layer address is to move a message from one computer on a circuit to the next computer on the same circuit.

Private addresses

Are used for servers and client computers that we want to keep private and are not intended for anyone on the Internet to find and send messages to. Private addresses are used inside a company and cannot be used on the public Internet. If a computer on the Internet tries to route a message to a computer with a private address, the routers at all Internet Service Providers will just throw the message away.

Public addresses

Can be found by anyone on the Internet and are used for servers intended to be public (e.g., web servers, email servers).

Subnet

Computers in the same part of the network (called a subnet) typically have similar addresses. 129.79.50.xx vs 129.79.51.xx.

DNS

Domain Name System

DHCP

Dynamic Host Configuration Protocol - Assigns IP Addresses automatically.

EIGRP

Enhanced Interior Gateway Routing Protocol (EIGRP) - Cisco

Ethernet Address Resolution

Ethernet address resolution is the process of translating an IP address into an Ethernet address. Ethernet addresses are only used on the same circuit or subnet, so any time a computer needs to resolve an Ethernet address, it is guaranteed to be on the same subnet. When your computer needs to resolve an IP address, it sends a request to all computers on its subnet using the Address Resolution Protocol (ARP) that says, "would the owner of this IP please send me its Ethernet address". The ARP request is a broadcast message that goes to all computers on the subnet. The one computer with that IP address sends an ARP response back that provides its Ethernet address.

Ethernet Address Assignment

Ethernet addresses are assigned by the manufacturer when the network card is built. They are burned into the hardware chip in the network card, so network managers don't need to worry about them.

network layer address

Every computer on the Internet has a network layer address, also called an IP address. Since IPv4 addresses are shorter that IPv6 addresses (4 bytes versus 16 bytes), my examples with use IPv4 addresses; it's faster to write 4 bytes than 16 bytes. An IPv4 address is written as four numbers, separated by periods (e.g., 129.192.4.12). The purpose of the IP address is to enable end-to-end routing; that is, to identify the best route from the sending computer to the destination computer.

IP Address Resolution

IP address resolution is the process of translating an application layer address (e.g., www.iu.edu) into an IP address. This is done by contacting the domain name server (DNS) provided by the network to which you are connected. For example, if you are connected to an Indiana University network, your laptop talks to the IU DNS. If you are connected to the Starbucks network, then your laptop talks to the Starbucks DNS. Your laptop sends a DNS request to the DNS server and the DNS server responds with the IP address your laptop requested.

Device Address

In order to route a message, every device on the Internet has a different, unique address so we can identify it from all the other devices.

Network addresses can be __________ or ___________.

Network addresses can be public or private.

Private IPv4 addresses start with the number _____ or _______

Private IPv4 addresses start with the number 10 (e.g., 10.1.12.99) or the numbers 192.168 (e.g., 192.168.1.101). Since they are private, any computer (or user) can use these addresses inside their network. If you check your IP address when you are at home, you'll probably see it has an address that starts with 192.168. Private addresses are great for security.

routing table.

Routers perform routing by using a routing table. A routing table lists all the circuits connected to a router and what addresses are reachable on that circuit. Most routers have only two circuits (one circuit into a building, and one circuit onto the campus backbone). In this case, the routing table is simple. The circuit into the building can reach all the addresses on one subnet (e.g., 129.79.55.x), and the campus backbone circuit goes to all other addresses). Routing tables for the campus core router (or routers on the Internet) are more complicated because they have more circuits.

RIP

Routing Information Protocol (RIP) is probably the most common protocol used by companies. It is a simple basic protocol that counts the number of hops to determine which route is the shortest.

Routing

Routing is the process of finding the best route (i.e., path) through the network from the sending computer to the final destination.

Routing

Routing is the process of finding the best route through network from the sender to the receiver. We call the steps in the route "hops." That is, moving a message from one router to the next router is one hop. Routing is done one hop at a time, with each router deciding the best next destination for the message. In other words, there is no overall master plan for the route; each router makes its own decision based on the circuits it knows (and sometimes considering how busy those circuits are).

Routing Protocols

Routing protocols are the languages that routers use to send messages to other routers about what networks they can reach. For example, if we add a building to our campus that has a new router with a new subnet, we configure that router, so it knows what subnet is inside the building. When we turn it on, it sends a message to the campus core router, saying it is a new router and can route messages to the new subnet. That way, the campus core router learns what subnets are reachable via which circuit.

hardware manufacturers

Self-Assessment Question 16 (Pick One) A client computer is assigned an Ethernet address by: - requesting one from a DNS server - software manufacturers - requesting one from a DHCP server - hardware manufacturers

requesting one from a DHCP server

Self-Assessment Question 17 (Pick One) A client computer is assigned an IP address by: - requesting one from a DNS server - software manufacturers - requesting one from a DHCP server - hardware manufacturers

The transmission of traffic status reports uses network capacity that could be used to transmit data

Self-Assessment Question 18 (Pick One) One weakness of routing protocols like OSPF and EIGRP that provide traffic reports is: - Routing does not reflect changing network conditions, such as circuits with heavy traffic - It is the job of the network manager, not the routers themselves, to maintain the routing tables. - They cannot be used with non-government networks - The transmission of traffic status reports uses network capacity that could be used to transmit data - It requires less processing by each router than RIP

Network Layer Functions

The network layer performs two main functions: addressing and routing.

IP

The protocol used at the network layer is IP, which stands for Internet Protocol.

Application Layer Address

The unique name that you and I use to identify the server. This is usually a meaningful text name, such as www.iu.edu or Kelley.iu.edu.

Types of Addresses

There are three different types of addresses, each of which is used for a different purpose: - application layer address (i.e. iu.edu.) - network layer address (aka IP Address) - data link layer address (?? aka mac address and ethernet address)

Versions of IP

There are two common versions in use, IPv4 and IPv6, and every device sold today speaks both versions to ensure that older devices that speak IPv4 can still function in the network.

Network Address Translation

Used when using your computer at home to send a web request to a public Web server. In this case, we have to translate the private address into a public address that can be used on the Internet. This is called Network Address Translation and is done by the router that connects your network at home to the Internet.

Assigning IP Addresses - Overview

When organizations want to connect to the Internet, they request addresses from an Internet Register who assigns them a unique number that is not in use. Some organizations get only one address and use private addresses inside their organization. Others get a range of addresses. For example, one range of addresses assigned to Indiana University is 129.79.x.x, which means IU controls all the addresses that start with these first two bytes and is free to assign the last two bytes as it wants.

BGP

border gateway protocol (BGP)

OSFP

open shortest path first (OSFP) - open standards