Lenuta Alboaie [email protected]

(1)

Computer Networks– http://www.info.uaic.ro/~computernetworks

Network Level

Lenuta Alboaie [email protected]

1

(2)

Content

• Network Level

• IPv4 Problem

• Context

• Characteristics

• Subnets

• Private Networks

• ICMP

• Address Resolution

• IPv6 - overview

• Details -> Future Course

(3)

Context

• Initial Situation

Before the Internet, only nodes from the same network could communicate with each other

3

Figure: Individual Network

(4)

Considerations

• Problem

– How to carry packages in a heterogeneous environment?

• Heterogeneity

• At lower levels: how to make the interconnection of a large number of independent

networks?

• At higher levels: how to provide support for a wide variety of applications?

• Scaling: how could we handle a large number of nodes and

(5)

Solution

• IP – Internet Protocol

5

(6)

Network Level

• IP protocol is used for autonomous systems (AS - Autonomous Systems) in order to interconnect

Figure: Internet - collection of interconnected networks

(7)

Network Level

• Role: offers connectionless services to transport datagrams from source to

destination; source and destination can be in different networks

• Each datagram is independent from the others

• This level does not guarantee the right transmission (loss, multiplier,…)

• +…Future Course

7

(8)

IP Protocol

• IPv4 Datagram

Data

(9)

IP Protocol

• IPv4 Datagram

• Common values for Version field are:

• 4 – IP Protocol (RFC 791)

(6 for IPv6 protocol (RFC 1883))

9

Data

Figure: IPv4 Datagram

Specify the

datagram header length

Specify the size of the entire datagram

(10)

IP Protocol

• IPv4 Datagram

• Type of service the field allows the host to communicate to the subnet (e.g.

routers) what type of service is desired

Data

Figure: Type of Service Field

(11)

IP Protocol

• IPv4 Datagram

• Identification field allows the host to

identify if the received segment is part of one datagram

11

Data

• Flags:

• DF (Don’t Fragment) – rooters can’t fragment the indicated datagram

• MF (More Fragments) – signals that the packet is a fragment, followed by others; last fragment has this bit 0

• Fragment offset field - represents the fragment placed in a datagram

(12)

IP Protocol

• IP Datagram

• Datagram's Fragmentation:

• Each fragment has the same structure as the IP datagram

• Reassembly of datagrams is performed by the receiver

• If a fragment of a datagram is lost, the datagram is

destroyed (an ICMP - Internet Control Message Protocol message is sent to the sender)

• Fragmentation mechanism has been used for some attacks - firewall piercing (a "special“ fragment is considered as

part of a connection already established, so that it can pass through a firewall)

(13)

IP Protocol

• IP Datagram

• Datagram Filtering:

• It is accomplished by a firewall: it allows access from the outside in the internal network,

according to some policy, certain types of packets ((used by certain protocols / services)

• Forestall a series of attacks regarding security

• The firewall can be software or hardware

• The firewall can function as a proxy or a gateway

13

(14)

IP Protocol

• Proxy- role and architecture:

– Indirect access to other networks (Internet) to hosts on the local network (via proxy)

– The proxy allows an Internet connection sharing – The proxy can be software or hardware

– May play roles such us: firewall or cache server

– Used to improve the performance (e.g., caching, flow

control), filtering messages, ensuring anonymity

(15)

IP Protocol

• IPv4 Datagram

• TTL (Time to Live) field specifies the lifetime of the package;

the number is

decremented by every router through which the packet passes

15

Data

(16)

IP Protocol

• IPv4 Datagram

• Protocol field specifies the protocol (from the superior level)

intended to process the datagram

Data

 1 ICMP (Internet Control Message Protocol)

 2 IGMP (Internet Group Message Protocol)

 6 TCP (Transmission Control Protocol)

 17 UDP (User Datagram Protocol)



(17)

IP Protocol

• IPv4 Datagram

• Header checksum field used for

detection; if an error occurs the datagram is

destroyed

17

Data

(18)

IP Protocol

• IPv4 Datagram

• Options Field

Data

Options Details

Security Mention if the datagram is a “secret” one

Strict source routing Show full path to go

Loose source routing Indicates a list of routers that should not be skipped

Record route Each rooter adds its own IP

(19)

IP Protocol

• IPv4 Datagram

• Source address and Destination

address fields indicate the

source address and the destination

address

19

Data

(20)

IP Protocol

• IPv4 Addresses

• Each IP address includes a network identifier(NetID) and a host identifier (HostID)

• Each network interface has a single IPv4 address

• An IPv4 address has a length of 32 bits

• Initially (RFC 791) there was a division into

network classes: A,B,C,D,E

(21)

IP Protocol

• IPv4 Addresses

21

[Computer Networks, 2003 Andrew S. Tanenbaum]

(22)

IP Protocol

• IPv4 Addresses

• Class A: 128 possible networks, 2²⁴hosts/network

• Class B: 2¹⁴ possible networks, 2¹⁶ hosts/network

• Class C: over 2 million networks, 255 hosts/network

• Network Identifier(NetID) is assigned by a central authority (NIC – Network Information Center)

• Host Identifier(HostID) is assigned locally by a network administrator

• Example: 85.122.23.145 – Class A (in decimal notation convention)

0101 0101 0111 1010 0001 0111 1001 0001

• For IPv6, hexadecimal representation is recommended

(23)

IP Protocol

• IPv4 Addresses

• An interface network has assigned a unique IP address

• A host can have multiple NICs, therefore it has multiple IP addresses

• The hosts of the same network have the same network identifier (the same NetID)

• Broadcast addresses have HostID’s bites equaled to 1

• The IP address in which all HostID’s bites are 0 is called a network address – refers to the hole network

• Example: 85.122.23.0 (network address for a host such us 85.122.23.145 and 85.122.23.1)

• 127.0.0.1 – loopback address (localhost)

23

(24)

IP Protocol

• IPv4 Addresses

• From the address space, some addresses are reserved: (RFC 1918):

• 0.0.0.0 – 0.255.255.255

• 10.0.0.0 – 10.255.255.255 (private addresses)

• 127.0.0.0 – 127.255.255.255 (loopback addresses)

• 172.16.0.0 - 172.31.255.255 (private addresses)

• 192.168.0.0 - 192.168.255.255 (private addresses)

• Private addresses : addresses that are not accessible to the outside (the "real“Internet), but only in the organization's intranet

(25)

Private Networks

• Aspects:

• The exponential growth of the hosts number

• Not all hosts offer resources available on the Internet Solution: NAT (Network Address Translation) – RFC

3022, 4008

• The private addresses can be reused (RFC 1918)

• It is based on replacing the private IP address with a public IP address (IP masquerading)

25

(26)

Private Networks

• Functionality:

Figure: NAT mechanism

• Routers normally ignore datagrams containing private addresses => private IP addresses can be used in the

organization's intranet

• Access to the outside (the "real“ Internet ) is achieved via a gate (mediating gateway) that rewrites the

source IP addresses / destination

(27)

IP Protocol

• Subnets using network masks

– It appeared as a solution to the problem of IP address space`s exhaustion

– Simplify Routing

– Subnets cannot be detected externally

27

Figure: A campus network

(28)

IP Protocol

• Subnets using network masks

• Division into subnetworks will be made via the network mask(netmask): NetID bits are 1, HostID bits are 0

• Subnet identifier (SubnetID) is generally used to group computers based on physical topology

Example. One way to create a subnet in a B network

(29)

IP Protocol

• Subnets using network masks

• Example:

– Let’s consider the IP address: 160.0.6.7 10100000 00000000 00000110 00000111

– Network mask: 255.255.252.0

11111111 11111111 11111100 00000000

– Network address: 160.0.4.0

10100000 00000000 00000100 00000000

Network address = network mask AND IP address

• Default subnet masks:

– 255.0.0.0 - Class A – 255.255.0.0 - Class B

• 255.255.255.0 - Class C

29

(30)

IP Protocol

• Convention mark:

x.x.x.x/m means that we apply an m bits mask to the IP specified x.x.x.x address

• Example:

• 10.0.0.0/12 – it applies a 12-bits mask to 10.0.0.0 address, we select possible values for the last 20 bits (=32-12)

• 85.122.16.0/20 – it applies a 20-bit mask to 85.122.16.0 address

(31)

Network Level

• Protocols

• ICMP (RFC 792)

• ARP (RFC 826)

• RARP (RFC 903)

• BOOTP (RFC 951,1048,1084)

• DHCP

• From IPv4 to IPv6

31

(32)

ICMP Protocol

• ICMP – Internet Control Message Protocol

• Used to exchange control messages

• Use IP

• ICMP messages are processed by the IP software, not by the user processes

• Messages types

Message Type Description

8 Echo Request Ask if a host is active 0 Echo Replay “Yes, I’m active”

3 Destination Unreachable The package can’t be delivered (e.g. DF is set) 5 Redirect The message is not correctly routed

11 Time Exceeded Time elapsed (TTL=0) <- (e.g. loop, congestions, low values for time)

(33)

Protocolul ICMP

• Used by:

• ping command (Packet Internet Gropher)

• traceroute command

• A package with TTL=1 (1 hop) is sent

• The first router ignores the packet and sends back an ICMP message “time-to-live exceeded”

• A package with TTL=2 is sent (2 hops)

• The second router ignores the packet and sends back an ICMP message “time-to-live exceeded”

• Repeat until it has received a response from the

destination or has reached the maximum number of hops

33

(34)

Address resolution

• IP addresses <-> hardware addresses (physical)

• The process of finding the hardware address of a host, knowing its IP address is called address resolution(address resolution) – ARP protocol (RFC 826)

• ARP –broadcast protocol (each host receives a request for a

physical address, and the answer is given by the one in question)

• The process of finding the IP address based on the hardware address is called reverse address resolution –RARP Protocol (RFC 903)

• Used to boot workstations without disks

• BOOTP (RFC 951,1048,1084)

• DHCP (Dynamic Host Configuration Protocol) RFC 2131,2132

(35)

IPv6

• Context:

– Issues in IPv4 addresses world:

• The exponential growth of the hosts` number

• Very large routing tables

• Complex configurations, more and more users (and increasing)

• Failure to ensure QoS

– Pressure from mobile operators

35

(36)

IPv6

• Objectives for a new protocol:

– Support for billions of hosts – Reducing routing tables

– Simplifying Protocol

– Support for mobile hosts

– Compatibility with the old IP

– Support for future developments of the Internet

– RFC 2460, 2553

(37)

IPv6

37

• 6 June 2012

(38)

IPv6

• Aspects:

– IPv6 addresses are 16 bytes in length - 2¹²⁸addresses

– Note: 16 hexadecimal numbers, 2 digits each, separated by “:”

• Example: 2001:0db8:0000:0000:0000:0000:1428:57ab

• If one or more groups of 4 digits is 0000, the zeros may be omitted and replaced (once) with "::“

• Example: 2001:0db8::1428:57ab

– To maintain compatibility, public IP addresses are considered a subset of IPv6 address space

– IPv4 addresses in IPv6 can be written as: 10.0.0.1 -> ::10.0.0.1 or 0:0:0:0:0:0:A00:1

(39)

IPv6

• ICMPv6

– ICMP provides functions (reporting data transmission, errors, etc.) plus:

• Neighbor Discovery(Neighbor Discovery Protocol – NDP) - Replaces the ARP

• Multicast listener discovery(Multicast Listener Discovery) – replaces IGMP (Internet Group Management Protocol)