The IP and Related Protocols

Transcription

1 The IP and Related Protocols IP, CIDR, NAT, ICMP, ARP, DHCP Summer 2013

2 Internet Service Internet provides an unreliable best effort, connectionless packet delivery system The service makes the earnesst attempt to deliver packets Delivery is not guaranteed Packets may be lost, duplicated, delayed or delivered out of sequence Packets are treated independently This service is defined by the Internet Protocol

3 Internet Protocol Essentially, IP defines: The basic unit of data transfer, Internet datagram Packets forwarded using destination host address Packets between same source and destination pair may take different paths A set of rules that characterize the best effort datagram delivery system No call setup at network layer No network-level concept of connection Routers keep no state about end-to-end connections

4 Datagram Format Vers HLen Service Type Total Datagram Length Identification Flags Fragment Offset Time to Live Protocol Header Checksum Source IP Address Destination IP Address IP Options (If Any) Padding Data... Vers = IPv4 HLen = Header Length

5 Datagram Type of Service Field Bit Precedence D T R Reserved The original intention of the TOS field was for a sending host to specify a preference for how the datagram would be handled as it travels through the Internet. In practice, the TOS field was not widely implemented in IPv4 Recently, the field is used in DiffServ to define a 6-bit Differentiated Services Code Point (DSCP) for packet classification

6 Precedence Field Precedence is a measure of the datagram importance. Precedence Hexadecimal Decimal Description 111 0xE0 224 Network Control 110 0xC0 192 Internetwork Control 101 0xA0 160 CRITIC/ECP 100 0x Flash Override 011 0x60 96 Flash 010 0x40 64 Immediate 001 0x20 32 Priority 000 0x00 0 Routine

7 D, T, and R Field D T R Delay Request Normal Low 0 1 Throughput Request Normal High 0 1 Reliability Request Normal High 0 1

8 Type of Service Router Processing Routers maintain a ToS value for each route in its routing table and uses it to select routes Routes learned through a protocol that does not support TOS are assigned a TOS of zero. If datagram ToS is set to a value other than zero, the router selects the route, with the best metric, whose ToS exactly matches the datagram ToS If no route qualifies for the specified ToS, the router drops the datagram and informs the source that destination is unreachable.

9 Fragmentation Datagrams have maximum length of 65,535 bytes Datagrams are encapsulated into physical frames Ideal case, IP datagram fits into a physical frame Often, fragmentation is required, depending on hardware Maximum Transmission Unit (MTU) IP expects gateways to handle datagrams of up to 576 bytes Reassemly of fragments takes place at the ultimate destination May lead to inefficiency When timers expire, the receiving machine drops the fragments

10 Fragmentation Control Identification 0 DF MF Fragment Offset Upon fragmentation, the total length refers to the size of the fragment, so it cannot be used to collect all fragments Identification: unique integer for every fragmented packet, all fragments share the same identification Fragment (DF) Flag and More Fragments (MF) Flag When set, DF forbids fragmenting of a datagramt When set, MF indicates that more fragments are to come Fragment offset offset in multiples of 8 bytes of given fragment in the packet Size of all fragments except for the last one must be divisible by 8

11 IP Fragmentation and Reassembly Length =4000 ID =x Fragflag =0 Offset = byte datagram MTU = 1500 bytes Length =1500 Length =1500 ID =x ID =x Fragflag =1 Fragflag =1 Offset =0 Offset =1480 Length =1060 ID =x Fragflag =0 Offset =2960 Resulting Fragments

12 Fragmentation and Reassembly IP specifies that datagram reassembly is done only at the destination Not on a hop-by-hop basis Requires maintaining timers for each datagram to avoid indefinite waiting If timer expires the datagram is dropped If any of the fragments is lost, the entire datagram is lost An Internet Control Message Protocol (ICMP) packet, containing the header of the lost packet, is sent to the host sender

13 Time to Live (TTL) The field specifies how long, in seconds, a datagram is allowed to live in the Internet Maximum time is estimated when a datagram is injected Each gateway decrements by one the TTL of each datagram it processes Each gateway decrements the TTL by the number of seconds the datagram waited inside the gateway When TTL reaches 0, the datagram is dropped Guarantees that datagram cannot travel around indefinitely Allows to cope with congestion

14 Protocol Number Indicates the higher level protocol to which IP should deliver the data 0 : reserved 1 : ICMP 2 : IGMP 3 : GGP 4 : IP encapsulation 5 : stream 6 : TCP 17 : UDP 89 : OSPF

15 Header Checksum It is a checksum for the header only and does not include the data Checksum is calculated as the 16-bit one s complement sum of all 16-bit words in the header For the purpose of this calculation, the checksum is assumed to be 0

16 IP Address IPv4 defines 4-byte (32 bits) address format IP address represents a network interface Routers, for example, are typically assigned multiple IP addresses Address space ~ = 4,294,967,296 addresses

17 IP Options Options are variable length An IP implementation is not required to be capable of generating options in the datagram it creates All implementations, however, are required to be able to process a datagram containing options There may be zero or more options in a datagram

18 Option Format Two option formats are possible A type byte alone A type byte, a length byte and one or more data bytes The type byte has the same structure in both cases The format of each option is dependent on the value of the option number found in the first byte of option

19 Datagram Options Type Byte Format Copy Option Class Option Number Copy : when set that option should be copied into all fragments Option Class: 0 : Datagram and Network Control 2 : Debugging and Measurement 1,3 : Unused

20 Option Number 0 : End of option list 1 : No operation This option may be used to align fields in the datagram 2 : Security 3 : Loose source routing 4 : Internet time stamp 7 : Record route 8 : Stream 9 : Strict source routing

21 Record Route Option Allows a source to create an empty list of IP addresses Each gateway that handles the datagram adds its IP address to the list Used for evaluation of routing paths

22 Source Route Option Provides a way for the sender to dictate a routing path Two forms of source routing Strict source routing The address specify the exact path the datagram should follow If the route cannot be satisfied, the gateway drops the datagram Loose source routing Specifies the path that must be followed by the datagram, but allows for multiple hops between succesive addresses in the list Useful for testing paths and reachability

23 Time Stamp Option Allow a source to create an empty list of IP addresses and time stamps (current time and date) The options are determined by a flag 0 : record time stamps only, omit IP addresses 1 : precede each time stamp with an IP address 3 : IP address is specified by the sender A gateway records a time stamp only if its address matches the next IP address

24 IP Addressing

25 IP Address Every device connected to the public Internet must be assigned a unique IP address. Typically addresses are assigned to Internet Service Providers (ISP) within region-based blocks, IP address can be used to identify the region or country from where Internet connection originates IP addresses can be assigned by an ISP statically (Static IP Address) or dynamically (Dynamic IP Address)

26 IP Address IPv4, defined by 4 bytes (32 bits) IP address represents a network interface Routers, for example, are typically assigned multiple IP addresses Address spaces ~ = 4,294,967,296 hosts

27 Classful IP Address Format Class A B C 0 NetID 10 NetID 110 NetID HostID HostID HostID P r i m a r y C l a s s e s D E 1110 Multicast Address 1111 Experimental Address 8 bits 8 bits 8 bits 8 bits

28 Class A Networks Class A networks are referred to as /8 networks, since they have an 8-bit network prefix /8 address block contains 2,147,483,648 individual addresses (2 31 ), or 50% of the total IPv4 address space: 4,294,967,296 (2 32 ) Max of 126 (2 7-2) /8 networks can be defined and are reserved Each /8 network supports 16,777,214 (2 24 2) hosts All-0 s and All-1 s numbers cannot be assigned to hosts

29 Class B Networks Class B networks are referred to as /16 networks, since they have an 16-bit network prefix /16 address block contains 1,073,741,824 individual addresses (2 30 ), or 25% of the total IPv4 address space: 4,294,967,296 (2 32 ) Max of 16,384 (2 14 ) /16 networks can be defined Each /16 network supports 65,534 (2 16 2) hosts All-0 s and All-1 s numbers cannot be assigned to hosts

30 Class C Networks Class C networks are referred to as /24 networks, since they have an 24-bit network prefix /24 address block contains 1,073,741,824 individual addresses (2 29 ), or 12.5% of the total IPv4 address space: 4,294,967,296 (2 32 ) Max of 2,097,152 (2 21 ) /24 networks can be defined Each /24 network supports 254 (2 8 2) hosts All-0 s and All-1 s numbers cannot be assigned to hosts

31 IP Addresses IP Address dotted decimal notation It divides the 32-bit IP address into 4 byte fields and specifies each byte independently as a decimal number with the fields separated by dots

32 Dotted Decimal Ranges Address Class A(/8 prefixes) B(/16 prefixes C(/24 prefixes) Dotted-Decimal Notation Ranges 1.xxx,xxx.xxx trough 126.xxx.xxx.xxx xxx.xxx through xxx.xxx xxx through xxx

33 Reserved IP Addresses Default route Loopback IP address Test IPC on local machine All bits are 0 in host number Denote this network All bits are 1 in host number Broadcast address in this network

34 Private Addresses Hosts directly connect to Internet must have globally unique public IP address For many organizations, most of the hosts do not need direct Internet connectivity Hosts typically access Internet services, e.g., Web and E- mail, through proxy servers and servers. Hosts within private Internets are assigned private addresses with the private address space. Public and private address spaces do not overlap Private addresses never duplicate public addresses.

35 Private Address Space Private address space is defined by three address blocks: /8 address block Class A network ID From to valid IP addresses, each with 24 host bits /12 address block It can be interpreted either as a block of 16 class B network IDs or as a 20-bit assignable address space (20 host bits) From to valid IP addresses: /16 address block it can be interpreted either as a block of 256 class C network IDs or as a 16-bit assignable address space (16 host bits) From to valid IP private addresses

36 Unforeseen Limitation of classful Addressing Addresses were allocated to organizations based their requests rather than actual need The decision to standardize on a 32-bit address space did not foresee a network of things Classes were easy to understand and implement but did not foster efficient allocation /24 is too small and /16 is too large Allocating a 16 to an organization that has several hundreds sites is wasteful and depletes the address space Allocating several /24 s increases the size of the routing table

37 Subnet Addresses Subnetting is introduced mainly to address depletion and routing table inflation and ease of management Three-Level Hierarchy The number of subnets must be a power of 2 10 NetID SubnetID HostID The subnet structure of a network is never visible outside the local network This limits considerably the expansion of the routing table

38 Sub-netting Subnet h Subnet h Subnet h

39 Default Mask and Subnet Mask Default Mask IP Address AND Network Address Subnet Mask IP Address AND Network Address

40 Logical Bitwise AND Operation Class B address: Binary: Subnet Mask: Binary: Network Address AND

41 Subnetting - Example A company is granted the a site Class C address: The number of 1s in the default mask is 24 The network configuration of the company requires six subnetworks Develop a satisfactory subnetting solution for this configuration

42 Subnetting Solution The number 6 is not a power of 2. The next number that is a power of 2 is 8 (2 3 ) 3 more bits are needed for the subnet mask The total number of 1s in the subnet mask is = 24 (original) + 3(added)

43 Subnetting Solution Subnet Mask In decimal dotted notation Number of subnets is 8. Number of available addresses in each subnet is (2 5 2) or 30 Address and address are reserved

44 Extended Network Address Internet routers use only the network number of the destination address to route traffic to a subnetted environment Routers within the subnetted environment use the extended network number to route traffic Extended Network Address NetID SubnetID HostID

45 Subnet Masks The extended-network prefix is identified by a subnet mask IP Address Mask A bit of the subnet mask is set to 1 if the corresponding bit in the IP address must be considered as part of the extended network prefix Extended Network Prefix Network Prefix IP Address:

46 Subnetting Prefix Advertisement Private Network Subnet ID Public Internet

47 Subnet Design Example An organization has been assigned the network number /16 Needs to create a set of subnets that supports up to 60 hosts on each subnet Step 1: Defining the Subnet Mask = 62, no room for expansion; = 126 Step 2: Defining subnet length and subnet numbers Step 3: Defining Hosts addresses for each subnet Step 4: Defining the broadcast address for each subnet

48 Internet Growth and Scalability By 1992, the exponential growth of the Internet the was causing concerns about the ability of the Internet's routing system to scale: The near-term exhaustion of the Class B network address space The rapid growth in the size of the global Internet's routing tables The eventual exhaustion of the 32-bit IPv4 address space The response to the two first immediate challenges: The concept of Supernetting or Classless Inter-Domain Routing (CIDR). The concept of Network Address Translation (NAT) The response to the 3rd problem, which is of long-term nature, is IPv6.

49 Classless Inter-Domain Routing CIDR

50 CIDR Main Features CIDR eliminates the traditional concept of Class A, Class B, and Class C network addresses. This enables the efficient allocation of the IPv4 address space CIDR supports route aggregation where a single routing table entry can represent the address space of up to thousands of traditional classful routes. Route aggregation reduces the amount of routing information in backbone routers, minimizes route flapping and eases the managers burden of updating external routing information. Without the rapid deployment of CIDR, the Internet would probably not be functioning today!

51 CIDR Network Prefix Concept CIDR replaces the traditional concept of address class with the generalized concept of a "network-prefix" Routers use the network-prefix, rather than the first 3 bits of the IP address, to determine the boundary between the network number and the host number. As a result, CIDR supports the deployment of arbitrarily sized networks rather Routing information is advertised with a variable bit mask, which represents the prefix-length. The prefix-length specifies the number of leftmost contiguous bits in the network-portion of the address Regardless of the class of the IP address, a network with 20 bits of network-number and 12 bits of host number is advertised with a 20-bit prefix length

52 CIDR Network Prefix /20 Bitwise Contiguous Address Blocks A /20 prefix can be assigned to a traditional Class A, Class B, or Class C network number. Each of the following /20 blocks represent 4,096 host addresses in /20, /20, and / / / /

53 Efficient Address Allocation Assume that an ISP has been assigned the address block /18. This block represents 16,384 (2 14 ) IP addresses or 64 /24s. Assume a client requires 800 host addresses Solution 1: ISP assigns the client a Class B ~64,700 addresses are wasted Solution 2: ISP assigns the client 4 individual Class Cs it introduces 4 new routes into the global Internet routing tables), Solution 3: ISP assigns the client the address block /22, a block of 1,024 (2 10 ) IP addresses (4 contiguous /24s). ISP's Block: /18 Client Block: /22 Class C #0: /24 Class C #1: /24 Class C #2: /24 Class C #3: /24

54 Classless Inter-Domain Routing (CIDR) CIDR was designed to address the ROADs problem No concept of address classes Prefixes are not restricted to /8, /16 and /24 Prefixes could be any length from 1 to 32 1<= masklength <=32 As a result, CIDR supports the deployment of arbitrarily sized networks rather than the standard 8-bit, 16-bit or 24- bit networks numbers Regardless of the class of the IP address, a network with 20 bits of network-number and 12 bits of host number is advertised with a 20-bit prefix length

55 Classful Subnetting and Address Allocation ISP owns the address block /16 and wants to allocate the /20 address block In a classful environment, it can only be cut into 16 equal-size segments / / / / / / / / / / / / / / / /24

56 Classless Subnetting and Address Allocation Address slices do not have to be of equal size Address block / /23 D /23 C A / /22 B

57 Controlling Routing Table Growth CIDR requires that the Internet be divided into addressing domains Within a domain, detailed information is available about all networks that reside in the domain Outside of an addressing domain, only the common network prefix is advertised This allows single routing table entry to specify a route to many individual network addresses

58 CIDR Controlling Routing Table Growth Internet Service Provider Internet / / / / / / / / / / / / / / / / / / / / / /24 Organization A Organization B Organization C Organization D

59 CIDR Forwarding Algorithm All routers must implement a consistent forwarding algorithm based on the "longest match" algorithm. A route with a longer extended-network-prefix describes a smaller set of destinations than the same route with a shorter extendednetwork-prefix. A route with a longer extended-network-prefix is said to be "more specific" while a route with a shorter extended-network-prefix is said to be "less specific." Routers must use the route with the longest matching network-prefix (most specific matching route) when forwarding traffic.

60 Classless Routing Longest Prefix Match -- Example Assume a packet's destination IP address is and there are three network prefixes in the routing table ( /24, /16, and /8) Destination = Route # /24 = * Route # /16 = Route # /8 = Router would select the route to /24. The /24 route is selected because its prefix has the greatest number of corresponding bits in the Destination IP address of the packet.

61 CIDR Summary CIDR allows efficient allocation of the IPv4 address space Divide old class A IP address into several reasonably sized IP prefixes / /24, /20, Aggregate several class C IP addresses into one reasonably sized prefix /24, / / /24, /24, /24, / /19 Routing uses the Longest Prefix Match The prefix /16 covers /20 The latter is more precise than the former

62 Network Address Translation NAT

63 Network Address Translation Combined with CIDR, NAT offers a short term solution to the problem of IP address depletion RFC-1631 NAT is designed to conserve IP addresses Use of private addresses, internally Long term solution is provided by IPv6

64 Network Address Translation Protocol NAT is a protocol that enables hosts on private networks to communicate with hosts on the Internet NAT is run on routers that connect private networks to the public Internet, NAT replaces IP addresses, and possibly port numbers, of IP datagrams at the boundary of a private network

65 Main uses of NAT Pooling of IP addresses Host migration support between network service providers IP address and port translation IP masquerading Load balancing of servers

66 Pooling of IP addresses Objective is to alleviate the public address shortage problem of corporate networks with large number of hosts and limited number of public addresses Corporate network is managed internally with a private address space NAT device, located at the boundary between the corporate network and the public Internet, manages a pool of public IP addresses NAT device selects a public IP address from the address pool, and binds it to the private address of the host

67 2: NAT router changes datagram source addr from , 3345 to , 5001, updates table Network Address Translation Address Pooling 2 NAT Device WAN side Addr LAN side Addr , , 3345 S: , 5001 D: , S: , 80 D: , : Reply arrives dest. address: , S: , 3345 D: , 80 1 S: , 80 D: , : host sends datagram to , : NAT router changes datagram dest addr from , 5001 to , 3345

68 Host Migration Between Network Service Providers Private Network ISP 1 Allocates Address Block /24 to Private Network Source = Destination = Source = Destination = Internet Service Provider 1 H1 Private address: Public address: NAT Device Source = Destination = ISP 2 Allocates address block /24 to private Internet network: Service Provider 2 Private Address Public Address ISP 2 Allocates Address Block /24 to Private Network

69 IP Address and Port Translation Address Port Translation Single public IP address is mapped to multiple hosts in a private network. Assign private addresses to the hosts of the corporate network NAT device modifies the port numbers for outgoing traffic

70 IP Address and Port Translation Private Network Internet Host 1 Source = Source port = 1500 Source = Source port = 4040 Private Address: NAT Device Host 2 Private Address: Source = Source Port = Source = Port = 5050 Private Address Public Address / / / /5050

71 Server Load Balancing A set of identical servers, accessible from a single IP address, are configured to provide similar service The objective is to balance the load among these servers Servers are assigned private addresses NAT device acts as a proxy for requests to the server orginating from the public network The NAT device substitutes the destination IP address of arriving packets to one of the private addresses for a server A typical strategy for balancing the load of the servers is to assign the addresses of the servers in a round-robin fashion.

72 NAT Limitation NAT use is problematic with: Protocols that require a separate back-channel Protocols that encrypt TCP headers Embed TCP address information Specifically use original IP for some security reason

73 Practical Objections Against NAT Port #s are meant to identify sockets Yet, NAT uses them to identify end hosts Makes it hard to run a server behind a NAT NAT Requests to on port Which host should get the request???

74 Services That Cause NAT to Fail! H.323, CUSeeMe, VDO Live video teleconferencing applications Xing Requires a back channel Rshell used to execute command on remote Unix machine back channel IRC Internet Relay Chat requires a back channel PPTP Peer-to-Peer Tunneling Protocol SQLNet2 Oracle Database Networking Services FTP Must be RFC-1631 compliant to work ICMP sometimes embeds the packed address info in the ICMP message IPSec used for many VPNs IKE Internet Key Exchange Protocol ESP IP Encapsulating Security Payload

75 Internet Protocol Routing PACKET FORWARDING

76 IP Routing An important function of the IP layer is routing Provides the basic mechanism for routers to interconnect different physical netowrks Usually, IP routers function with partial information Two types of routing are possible Direct Routing Indirect Routing

77 IP Routing Direct Forwarding If the destination host is attached to a physical network to which the source host is also attached, a datagram is sent directly The datagram is encapsulated into the physical frame and delivered directly to the destination host

78 IP Routing Indirect Routing Indirect routing occurs when the destination host is not attached directly to the source host One or more IP routers are needed to reach the destination host Address of the first router is called indirect route in the context of the IP routing algorithm The address of the first router is the only information needed by the source

79 Indirect Routing over Subnets In some cases, multiple subnets may be defined in the same network If the destination host is on a different subnet than the source host, indirect routing is used (though hosts may be attached directly to each other) Therefore, there is a need for a router to forward traffic between subnets

80 IP Routing Table Whenever the IP routing software in a host or a gateway needs to transmit a datagram, it consults a routing table Both hosts and gateways have routing tables typically containing The IP address of each possible destination network, and The IP address of the next gateway along the path to the destination network

81 Routing Algorithm Route ( Datagram, Routing Table ) Extract destination Netid If Netid is of a direct neighbor, then send datagram directly Else if there is a route specified by host, route datagram as specified in the table (based on the information in the frame) Else if Netid is in the routing table, then route datagram as specified in the table Else if a default route has been specified, route datagram to the default gateway Else declare a routing error

82 Routing with Masks Route (Datagram, Routing Table, Mask) If (destination address & mask) = (my address & mask), then send the datagram directly Longest Prefix Match Else route the datagram to a next hop (as specified by the routing table)

83 Longest Prefix Matching Routes in routing table 1 st : /0 2 nd : /16 3 rd : /20 4 st : /8 2 nd prefix covers 3 rd prefix 3 rd prefix is more specific than 2 nd prefix For destination IP address: , choose eth , choose eth2 Longest prefix matching Choose route of more specific matching prefix Otherwise, it choose default route eth0 (default) eth1 eth2 eth0

84 Address Resolution Protocol ARP

85 IP Addresses to Hardware Addresses Mapping IP Addresses are not recognized by hardware Given the IP address of a host, how do we obtain the corresponding hardware address? This process of is referred to as the Address Resolution Local Process

86 Address Resolution Protocol The Address Resolution Protocol is used by a sending host which seeks to resolve the IP address of the destination into the corresponding Ethernet address. Ethernet address will carried into the Ethernet frame which encapsulates the IP datagram ARP is a broadcast protocol Every host on the network receives the request. Each host checks the request against its IP address, upon receipt of the request Only the sought after station responds.

87 ARP Caching ARP request is not necessary every time an IP datagram is sent Hosts remember the hardware addresses of each other. Caching To further improve performance, ARP protocol specifies that the receiving host should also remember the IP and hardware addresses of the sending host

88 ARP conversation ARP_Rqst( , MAC:?) Drop ARP_Rqst() ARP_Rply( , 87:A2:15:35:02:C3

89 Dynamic Host Configuration Protocol DHCP

90 Who manages the IP address? Multiple regional routing registries. North and South America American Registry for Internet Number (ARIN) Europe Reseaux IP Europeans (RIPE) Asia: Asia Pacific Network Information Center (APNIC) How do you obtain IP addresses? ISP buys from registries or from their provider Buy IP addresses from your provider Customer keeps IP address even after switching to another provider Rent IP address from your provider Return the IP address to your provider when you switch to another provider

91 How to Assign IP Addresses to Interfaces? Once an organization has obtained a block of addresses, it can assign individual addresses to the host and router interface in its organization Router IP Address Assignment System administrators manually configure the IP addresses into the routers Remotely, if necessary, using network management tools Host IP Address Allocation Manually, typically using a configuration file Dynamic Host Configuration Protocol (DHCP)

92 DHCP Dynamic Host Configuration Protocol (DHCP) DHCP is the preferred mechanism for dynamic assignment of IP addresses Designed in 1993, as an extension of BOOTP DHCP can interoperate with BOOTP clients Uses port numbers as BOOTP DHCP Extensions: Support for temporary allocation ( leases ) of IP addresses DHCP client can acquire all IP configuration parameters

93 DHCP Packet Format op (1 byte) htype (1 byte) hlen (1 byte) hops (1 byte) xid (4 bytes) secs (2 bytes) flags (2 bytes) ciaddr (4 bytes) yiaddr (4 bytes) siaddr (4 bytes) giaddr (4 bytes) chaddr (16 bytes) sname (64 bytes) file (128 bytes) options (variable)

94 DHCP Packet Fields op Message Type 1 = BOOTREQUEST: Client to server 2 = BOOTREPLY: Server to client htype Hardware Address Type 1 = 10Mb Ethernet hlen Hardware Address Length (in bytes) 6 (bytes) for 10Mb Ethernet. hops Hops taken so far Client sets to 0. Optionally used by relay agents when booting via relay agent. xid Transaction Id. Unique number to associate messages. Random number chosen by the client. secs Number of seconds elapsed since client began address acquisition/renewal Filled in by the client.

95 DHCP Packet Fields Flags: 16 bits B: Broadcast Flag (1 Bit): 1 = Broadcast 0 = Unicast Must Be Zero (15 Bits): For future expansion. zero! ciaddr Client IP Address Only filled in if client is in BOUND, RENEW or REBINDING states. yiaddr Your IP Address The IP Address that the server gives to the client.

96 DHCP Packet Fields siaddr Server IP Address Address of next server to use. Set by server in DHCPOFFER and DHCPACK. giaddr Gateway/ Relay Agent IP Address. Used if indirect connection to the DHCP Server. chaddr Client Hardware Address The Ethernet/MAC Address of the client. sname Server Name Optional server name. Null terminated string. file Boot File Name Null terminated string. Ignore for lab. options Various optional fields.

97 Options Fields Message Type. Present in most real implementations of DHCP. Makes packet type easier to identify. 1 = DHCPDISCOVER 2 = DHCPOFFER 3 = DHCPREQUEST 4 = DHCPACK 5 = DHCPNAK

98 DHCP Basic Operations If a client does not have an IP Address it must broadcast packets. Broadcast (B) Flag should be set. DHCPDISCOVER Client may not know which DHCP servers exist. DHCPREQUEST Broadcast in response to one or more DHCPOFFERs. This implicitly rejects other DHCPOFFERS. When Rebooting the client does not know if the address is now allocated to another node. Therefore it should broadcast. If the client broadcasts the server should respond with a broadcast. If the client has an IP Address it can unicast to the server. Broadcast (B) Flag should be unset Extending the lease. Server should respond with unicast

99 DHCP Operation DHCP Client 00:a0:24:71:e4:44 DHCP Server 1 DCHP DISCOVER DHCPDISCOVER Sent to DHCP Server 2 DHCP Client 00:a0:24:71:e4:44 DHCPOFFER DHCP Server 1 DCHP OFFER DHCPOFFER DHCP Server 2

100 DHCP Operation DHCP Client 00:a0:24:71:e4:44 DHCP Server DCHP DISCOVER DHCPREQUEST DHCPACK Upon receipt of the DHCPACK, the DHCP client can start to use the IP address DHCP Server DHCP Client 00:a0:24:71:e4:44 DHCP Server Renewing a Lease: Sent when 50% of lease has expired. The DHCP server can refuse lease extension by sending a DHCPNACK DHCPREQUEST DHCPACK DHCP Server

101 DHCP Operation DHCP Client 00:a0:24:71:e4:44 DHCP Server DCHP RELEASE: The DHCP client releases its IP address DHCPRELEASE DHCP Server

102 Internet Control Message Protocol ICMP

103 Internet Control Message Protocol ICMP is a required companion of IP Allows hosts to interact with gateways Provides feedback from gateways to hosts about problems Destination unreachable Time exceeded Parameter problem Source quench Redirect message to update route to a shorter one Echo and echo reply Timestamp request and reply for sampling delays

104 ICMP Messages ICMP message is encapsulated in an IP datagram IP Header ICMP Message IP Datagram

105 ICMP Message Format There are 15 different types of messages Some types of messages use different code values Checksum covers the entire ICMP message Message Type Code Checksum Field Type Based Content

106 ICMP Message Types PING Echo Packet InterNet Groper (PING), tests whever another host is reachable by sending ICMP Echo Request message to the host, expecting ICMP Echo Reply from that host Type Code Description Query Error 0 0 Ping Echo Reply 8 0 Ping Echo Request

107 Destination Reachability Type Code Description Query Error 3 Destination Unreachable 0 Network Unreachable 1 Host Unreachable 2 Protocol Unreachable 3 Port Unreachable 4 Fragmentation Needed, but Forbidden 5 Source Route Failed 6 Destination Network Unknown 7 Destination Host Unknown

108 Destination Reachability Cont. Type Code Description Query Error 3 Destination Unreachable 8 Source Host Isolated (obsolete) 9 Destination Network Administratively Prohibited 10 Destination Host Administratively Prohibited 11 Network Unreachable for TOS 12 Host Unreachable for TOS 13 Communication Prohibited by Filtering 14 Host Precedence Violation 15 Precedence Cutoff in Effect

109 Elementary Flow Control The source quench message contains an error that may be sent by a host or a router, when it receives data faster than it can handle A system is not required to send a source quench message, even if it runs out of buffers and throws datagrams away Type Code Description Query Error 4 0 Source Quench

110 Redirect Messages Type Code Description Query Error 5 Redirect 0 Redirect for Network 1 Redirect for Host 2 Redirect for TOS and Network 3 Redirect for TOS and Host

111 ICMP Messages Router Advertisement Type Code Description Query Error 9 0 Router Advertisement 10 0 Router Solicitation Time Expiration Type Code Description Query Error 11 Time Exceeded 0 Time_To_Live is Zero During Transmit 1 Time_To_Live is Zero During Reassembly

112 Parameter Errors Type Code Description Query Error 12 Parameter Error 0 Bad IP Header (catchall error) 1 Required Option Missing

113 Time stamps Type Code Description Query Error 13 0 Timestamp Request 14 0 Timestamp Reply

114 Information Type Code Description Query Error 15 0 Information Request (obsolete) 16 0 Information Reply (obsolete)

115 Address Mask Type Code Description Query Error 17 0 Address Mask Request 18 0 Address Mask Reply

116 ICMP Rules ICMP error messages are never generated in response to an ICMP error message to avoid recursive error generation When an ICMP error message is sent, the message always contains the IP header and the first 8 bytes of the IP datagram that caused the error Allows the ICMP receiver to uniquely idetify the protocol, and the protocol port number associated with the message

117 ICMP Rules ICMP error message is never generated for a datagram destined to an IP broadcast address, or a datagram sent as a link-layer broadcast, or a fragment other than the first, or a datagram whose source address does not define a single host (zero address, loopback address, broadcast address or multicast address) These rules are meant to prevent broadcast storms

118 Conclusion Internet Protocol and Best Effort Service IP Datagram Format and Fields IP Addressing Subnetting CIDR Network Address Translation NAT IP Datagram Forwarding Longest Prefix Match Address Resolution Protocol ARP Host Configuration DHCP Internet Error and Message Control ICMP