IP: (Internet Protocol) IP - 1

Transcription

1 TCP/IP Family of Protocols IP: (Internet Protocol) IP - 1

2 A suite of protocols What is TCP/IP? Rules for sending and receiving data across networks Addressing Management IP - 2

3 What is TCP/IP? TCP/IP stands for Transmission Control Protocol/ Internet Protocol. TCP/IP is a collection of protocols, or rules, that govern the way data travels from one machine to another across networks. The Internet is based on TCP/IP. IP - 3

4 What is TCP/IP? TCP/IP has two major components: TCP and IP IP: envelopes and addresses the data enables the network to read the envelope and forward the data to its destination (routing) defines how much data can fit in a single "envelope" (a packet) IP - 4

5 What is IP? The relationship between data, IP, and networks is often compared to the relationship between a letter, its addressed envelope, and the postal system. Best-effort delivery: -Packets are lost -packets are delivered out of order -duplicate copies of a packet are delivered -packets can be delayed for a long time IP - 5

6 What is TCP? TCP: breaks data up into packets that the network can handle efficiently (e.g. letter with 10 pages, only one page per envelope) verifies that all the packets arrive at their destination "reassembles" the data TCP can be compared to reliable postal services e.g., registered mail (Einschreiben), where the sender receives an answer on delivery. IP - 6

7 What is TCP/IP? In Summary: TCP/IP is a suite, or family, of protocols that govern the way data is transmitted across networks. TCP/IP protocols work together to break the data into small pieces that can be efficiently handled by the network, communicate the destination of the data to the network, verify the receipt of the data on the other end of the transmission, and reconstruct the data in its original form. IP - 7

8 TCP/IP vs OSI Layers APPLICATIONS 5 7. Session, Presentation and Application layers TCP / UDP IP LAN 4. Transport Layer 3. Network Layer 1 2. Data LL and PHY layers UDP : User Datagram Protocol is a rapid transmission protocol that uses IP packets to deliver data with no reliability features like connections and ACKs. The strength of UDP is speed, not reliability. IP - 8

9 TCP/IP Protocol Stack F T P T E L N E T N F S S M T P H T T P.. R T P T F T P D N S.. F T P T E L N E T N F S S M T P H T T P.. R T P T F T P D N S.. TCP UDP TCP UDP IPv4 ICMP ARP IPv6 ICMPv6 ND Ethernet, Token-Ring, x, ATM, X.25,.. Ethernet, Token-Ring, x, ATM, X.25,.. ICMP: Internet Control Message Protocol (RFC 792) ARP: Address Resolution Protocol (RFC 826) ICMPv6: ICMP for IPv6 (RFC 4443) ND: Neighbor Discovery Protocol for IPv6 (RFC 4861) IP - 9

10 Application Layer Protocols FTP File Transfer Protocol allows the transfer of copies of files between one node and another. FTP is not hardware-dependent so its services can function just about anywhere. Using this utility to copy data is typically referred to as "FTPing" a file. FTP usually uses TCP as a transport protocol. NFS Network Filing System was developed by Sun Microsystems Inc. It provides shared access to files in a very transparent and integrated way. Older versions of NFS uses both UDP and TCP as a transport protocol. The latest recommends to use TCP. TELNET Remote Terminal Emulation allows users to communicate with diverse hosts. The TELNET application provides terminal-type access from one network computer to another. SMTP Simple Mail Transfer Protocol is the middle-man that uses TCP to move data around from one internetwork host to another. Applications run on both hosts using SMTP. IP - 10

11 Other Protocols ICMP Internet Control Message Protocol offers flow control and error-detection to the unreliable delivery method of IP. It provides a facility for routers and gateways on the network to communicate with a source if there is a problem. It also provides a mechanism for determining if a destination cannot be reached. (IP layer) ARP Address Resolution Protocol is a special protocol to allow TCP/IP to interact in environments such as Ethernet. ARP maps TCP/IP addresses to Ethernet Data Link layer addresses. - more details in the exercise IP - 11

12 The Hour Glass www phone... SMTP HTTP RTP... TCP UDP... IP www phone... SMTP HTTP RTP... TCP UDP... IP + mcast + Qos +... ethernet PPP... CSMA async sonet... copper fiber radio ethernet PPP... CSMA async sonet... copper fiber radio [Source: Steve Deering - Cisco: 51st IETF; London, England, Aug. 5-10, 2001] IP - 12

13 IP Frame (Version 4: IPv4) 32 bits Version Internet Header Length Service Type Total Length ID Flags Fragment Offset Time to Live Protocol Header Checksum Source Address Destination Address Options Padding DATA IP - 13

14 IP-PDU Version (4 bit): IP-Version, 4 Internet Header Length (4 bit): total header length in multiples of 32 bit, e.g. 5 means the header consists of 5 x 32 bit = 160 bit or 20 byte (min. length) Type of Service (8 bit): can be used to set priorities for packets, see RFC 3168 Bits 0-5: DSCP (Differentiated Services Code Point), Bits 6-7: ECN (Explicit Congestion Notification / IP Flow Control) Total Length (16 bit): total length of packet incl. header in byte What s max length? ID (16 bit): ID and the following 2 fields (Flags and Fragment Offset) control the reassembly of fragmented IP-Packets. Unique ID of a datagram. This together with the 'Source Address is used to identify fragments for reassembly. Flags (3 bit): Bit 0: reserved, must be 0 Bit 1: 0/1 may/may not be fragmented Bit 2: 0/1 last fragment/more fragments to follow IP - 14

15 IP-PDU Fragment Offset (13 bit): a number specifying the position of the fragment, offset is given in steps of 64 bit/ 8 byte (first fragment has value 0) Time to Live (8 bit): specifies the time to live of a packet. A value of zero means the packet will be deleted. Every router on the path decrements the counter by one. This is used to avoid packets circulating in the network. The standard of 1981 defined it as the time in seconds that a packet spent at a station with a min. of one. Today it is used as a hop-count. Protocol (8 bit): specifies the contained protocol, e.g. for a TCP packet: 0x06, for a UDP packet: 0x11 (defined by IANA). Header Checksum (16 bit): checks the header Source Address (32 bit): IP Address Destination Address (32 bit): IP Address Options Padding (to achieve multiples of 32 bit) DATA - any guess about the first byte of Data? IP - 15

16 What is an IP Address? If you want to connect to another computer, transfer files to or from another computer, or send an message, you first need to know where the other computer is - you need the computer's "address." A way to identify machines on the Internet (more precisely: network interfaces) IP - 16

17 Features of an IP Address A globally unique number No two machines can have the same IP number Standardized All machines connected to the Internet agree to use the same scheme for establishing an address IP addresses are also referred to as IP numbers or Internet addresses. There are two versions of Internet Protocol: IPv4 and IPv6 IP - 17

18 What is an IPv4 Address? An IPv4 address consists of four sections separated by periods (dot-decimal notation). IP address: Network Host Each section contains a number ranging from 0 to 255. These 4 sections represent both the machine itself, Host and the Network that the Host is on. How many bits does an IPv4 address have? IP - 18

19 Who assigns an IPv4 Address? IP address blocks with a fixed network portion of the IP addresses are allocated to Internet Service Providers (ISPs) by the Regional Internet Registries (RIRs), who receive the address blocks from the Internet Assigned Numbers Authority (IANA). ISPs then assign the host portion of the IP address to the machines on the networks that they operate. IP - 19

20 The Regional Internet Registries (RIRs) were established to assume this regional allocation and management role in cooperation with IANA. Today, there are five RIRs - APNIC,ARIN, RIPE NCC, Lacnic, and AFRINIC. Source: IP - 20

21 Format of IP Addresses Globally unique (or made to seem that way) Hierarchical: network + host - Aggregating addresses saves memory in routers, simplifies routing Originally, routing prefix is embedded in an address: -There are 5 classes of IP addresses: Class A, Class B, Class C, Class D (multicasting), and Class E (reserved for future use). Now, routing info on CIDR blocks, address + prefix length - E.g., /16 IP - 21

22 IPv4 Address Classes The diagram below compares Class A, Class B and Class C IP addresses. The blue numbers represent the network and the red numbers represent hosts on the network. Therefore, a Class A network can support many more hosts than a Class C network. Class A Class B Class C Network Host IP - 22

23 Subnet Mask & IP Sub Networking Network part of IP address is given as a Subnet Mask Subnet Mask: is usually specified either by four sections separated by dots, e.g , or by a slash followed by the number of 1 -bits in the mask, e.g. /16 E.g or /16 IP sub networking: The way sub-networking operates is to borrow one or more of the available host bits and then make interfaces locally interpret these borrowed bits as part of the network bits /17 What is the subnet mask? Is a valid subnet mask? IP address calculations: e.g.: IP - 23

24 Configuration of IPv4 Address Manual configuration Automatic DHCP (Dynamic Host Configuration Protocols) Operations of DHCP (simplified version), more details in RFC IP - 24

25 Host A IP Forwarding Host B R 1 R 2 Needs for address translation Host A -> HTTP Request to ComNets website (Host B) - L3 translation: DNS e.g L2 Translation: ARP IP address to MAC address: (IP address of R1) - 01:23:45:67:89:ab (Ethernet MAC address of R1) IP vs MAC address (size, scope & which layer, address assignment)? Forwarding /routing (by R1 & R2) Fragmentation (Ethernet MTU = 1500 bytes, PPP MTU = 576 bytes) IP - 25

26 IP Configurations Two approaches for calculating the routing tables: Static Routing Dynamic Routing - Routes are calculated by a Routing Protocol IP - 26

27 Routing Protocols (Interdomain vs Intradomain) Autonomous System is a region of the Internet that is administered by a single entity and that has a unified routing policy Intradomain routing protocols IGP Protocols (e.g. OSPF, RIP,..) Routing is done based on metrics (hop counts, delays, etc) Routing domain is one autonomous system Interdomain routing protocols EGP Protocols (e.g. BGP) Routing is done based on policies Routing domain is the entire Internet 27

28 Problems of IPv4 Some deficiencies unsuitable for fast-growing Internet: Limited and inefficient use of address space two-level address structure (netid and hostid) categorized into 5 classes A, B, C, D and E (( ) reserved for future use), F (extension) cannot accommodate new Internet applications, such as real-time audio and video transmission maximum delay and reservation of resources required need more complex addressing and routing capabilities must accommodate encryption and authentication of data IP - 28

29 [Source: IP - 29

30 One Solution - NAT NAT : Network Address Translation Assign private addresses to the internal systems Router translates the addresses Global IP address Space NAT NAT Private Address Space Private Address Space IP - 30

31 How does NAT work? NAT is often built into routers. It receives each packet from the internal private network and modifies the IP header to match the global IP address of the router, before it is transmitted out into the Internet. The router stores the internal IP address, destination IP address and port number in a routing table so when a request is returned on the same port, the NAT can match the internal IP address that originated the request, and then modify the IP header to match that of the internal address. IP - 31

32 Why not NAT? NAT breaks end-to-end communication Routers monitor the communication Routers change the data NAT breaks Bi-directional communication Hosts with a global address cannot initiate the communication to the hosts with private addresses Advantages of NAT? IP - 32

33 Summary: NAT NAT (Network Address Translator) popular on Dial-ups and VPN networks saves IPv4 addresses loss of the end-to-end model asymmetric identifier/communication model IP - 33

34 What is IPv6? IPv6 is short for "Internet Protocol Version 6". IPv6 is the "next generation" protocol designed by the IETF to replace the current version Internet Protocol, IP Version 4 ("IPv4"). Today's Internet mostly uses IPv4, which was defined in 1981 (RFC 791). IPv4 has been remarkably resilient in spite of its age, but it is beginning to have problems. Most importantly, there is a growing shortage of IPv4 addresses, which are needed by all new machines added to the Internet. IPv6 fixes a number of problems in IPv4, such as the limited number of available IPv4 addresses. It also adds many improvements to IPv4 in areas such as routing and network auto-configuration. IPv6 is expected to gradually replace IPv4, with the two coexisting for a number of years during a transition period. IP - 34

35 IPv6 vs IPv4 Format of Address 128-bit IPv6 address = 64-bit prefix + 64-bit Interface ID (IID) Image source: Indeterminant (Wikipeida) 128-bit IPv6 address : each 16 bit values (four hex digits) separated by colons, one sequence of allzero 16-bit values can be replaced by a double colon indicating a longer sequence of zeros IP - 35

36 IPv6 Address Configuration 64-bit network Prefix 64-bit Suffix / Interface Identifier (IID) Stateless Address Auto Configuration EUI 64-bit Stateful Address Configuration DHCPv6 RFC 3315 EUI 64-bit : Appendix A RFC 4291: IP Version 6 Addressing Architecture Example: MAC address: 00:1D:BA:06:37:64 converts to 64-bit EUI address IID: 00:1D:BA:FF:FE:06:37:64 - Based on MAC address Network Prefix: 2001:db8:1:2::/64 Full IPv6 Address : 2001:db8:1:2:021D:BAFF:FE06: IP - 36

37 IPv6 Neighbor Discovery Protocol RFC 4861 ND: How hosts and routers interact with each other on the same link IPv6 is the format - ND is the brain One-hop routing protocol defined in RFC4861 Finding Routers Router Solicitation / Router Advertisement Finding Neighbors Neighbor Solicitation / Neighbor Acknowledgement Detecting Duplicate Addresses using NS/NA Address resolution using NS/NA Neighbor Unreachability Detection using NS/NA ND: Neighbor Discovery NS: Neighbor Solicitation NA: Neighbor Advertisement DAD: Duplicate Address Detection IP - 37

38 IPv6 Addresses in Hexadecimal Notation Long form Abbreviated form Explanation 2001:DB8:0.0:8:800:200C.417A 2001:DB8::8:800:200C:417A A unicast address FF01:0:0:0:0:0:0:101 FF01::101 A multicast address 0:0:0:0:0:0:0:1 ::1 Loopback address FE80:0:0:0: FE80:: A Link Local address IP - 38

39 IPv4 IPv6 header Ver4 Identification IHL Type of service Total Length Flags Fragmentation Offset Time To Live Protocol Header Checksum Source Address Destination Address Options Italics: removed in IPv6 IP - 39

40 IPv6 header = > 340 * RFC IP - 40

41 IPv6 Version - This field is the only field that kept the same meaning from IPv4 to IPv6. The 4-bit version field contains the number 6. It is the same size as the IPv4 version field, which contains the number 4. However, this field is usually not used to distinguish between IPv4 and IPv6 packets. The protocol type field present in the layer 2 envelope is used for that. Traffic Class - The 8-bit priority field used by the originating node or routers to identify different classes and priorities of an IPv6 packet. Flow Label - The 20-bit Flow Label field is used to label a set of packets that belong to the same flow. E.g., put packets to the same path for real time applications Payload Length - This 16-bit field is similar to the IPv4 Total Length Field, except that with IPv6 the Payload Length field is the length of the data carried after the header, whereas with IPv4 the Total Length Field included the header. IP - 41

42 IPv6 Next Header - This 8-bit field reflects the new organization of IP packets with IPv6. In IPv4, the IP header is always immediately followed by the transport protocol data. (E.g. UDP or TCP) and defined by the IPv4 Protocol Type field. With IPv6 we can have the same structure and set the next protocol type to be UDP (16) or TCP (6), or we can interleave Extension Headers between the IP and TCP/UDP payload. The next header type will then be set to the type of the first Extension Header. Hop Limit - This 8-bit field defines by number a count of the maximum hops that a packet can remain in the network before it is destroyed. With the IPv4 TTL field this was expressed in seconds and was typically a theoretical value and not very easy to estimate and was then practically used as a counter as well. Source Address - This 128-bit field contains the Source IPv6 address of the packet. Destination Address - This 128-bit field contains the Destination IPv6 address of the packet. IP - 42

43 IPv6 Extension Headers The current IPv6 specification defines 6 extension headers Hop-by-Hop Options Header Routing Header Fragment Header Authentication Header Encrypted Security Payload Destination Options Header IP - 43

44 University Bremen 8 blocks of 16 bit (hexadecimal notation -> 4 digits, leading zeros can be omitted, e.g. 2001:db8:0:8d3:0:8a2e:70:7344, sevaral blocks of zeros can be abbreviated to ::, but only once!!!, why?) University of Bremen address range: 2001:638:708::/48 FB1: 2001:638:708:1::/52 ComNets: 2001:638:708:1155::/ :638:708:1156::/64

45 Transition Mechanisms (Header Translation) RFC 2893 (Transition Toolbox) Dual Stack (Dual IP layer) IPv4 & IPv6 Configured/Automatic Tunneling of IPv6 over IPv4 IPv4-compatible IPv6 addresses (embedded IPv4 addresses) IP - 45

46 IPv4 to IPv6 Transition dual stack router IPv6 Workstation IPv6 IPv4 IPv6 dual stack router IPv6 Server IP - 46

47 Tunneling IPv6-over-IPv4 tunneling IPv6 packets are encapsulated within IPv4 Overhead? Configured Tunneling Tunnel endpoint address is determined by configuration information on the encapsulating node (point-to-point tunnels) Automatic Tunneling IPv4-compatible IPv6 addresses (embedded IPv4 addresses) Tunnel endpoint address is determined from the embedded IPv4 destination address IP - 47

48 IPv4 to IPv6 implications DNS: Domain Naming Service New record types to differentiate type of address Tunnels: Router-to-Router Host-to-Router Host-to-Host Router-to-Host Fragmentation IPv4 layer fragmentation should be avoided IPv4 path MTU Discovery Protocol IP - 48

49 IPv6 packet encapsulation in IPv4 IPv6 Header Transport Layer Header Data IPv4 Header IPv6 Header Transport Layer Header Data IP - 49

50 Industry Activities IPv6 Forum Product Presentations Product Demonstrations IETF Working groups 6bone 6REN 6TAP MWIF (Mobile Wireless Internet Forum) 3GPP (3 rd Generation Partnership Project, UMTS) Interoperability Testing Annually at Connectathon Annually at University of New Hampshire IPv6 testing at ilabs at Networld+Interop IPv6 Forum See also for information on IPv6 penetration IP - 50