Layered Architecture

Size: px

Start display at page:

Download "Layered Architecture"

Leo Ryan
5 years ago
Views:

1 1 Layered Architecture Required reading: Kurose 1.7 CSE 4213, Fall 2006 Instructor: N. Vlajic Protocols and Standards 2 Entity any device capable of sending and receiving information over the Internet (i.e. computer network) Protocol set of rules that governs data communication between two or more entities key elements: syntax, semantics, timing (1) Protocol Syntax format of protocol data units, i.e. number and size of fields in a packet and how these fields are delineated e.g. packet = 4 bytes header + 12 bytes data (2) Protocol Semantics meaning associated with bits carried in each control field (3) Protocol Timing when data should be sent and how fast Agreements about message formats and actions to to be be taken when receiving a message are called protocol specifications. 1

Protocols and Standards (cont.) 3 Example [ TCP-protocol syntax, semantics, timing ] Protocol Syntax: Protocol Timing: connection establishment Protocol Semantics: e.g. HLEN field indicates the length of the TCP header in 32-bit words (i.

Protocols and Standards (cont.) 4 Standard provide guidelines to manufacturers, vendors, service providers, etc.

2 Protocols and Standards (cont.) 3 Example [ TCP-protocol syntax, semantics, timing ] Protocol Syntax: Protocol Timing: connection establishment Protocol Semantics: e.g. HLEN field indicates the length of the TCP header in 32-bit words (i.e. # of 4-byte segments in the header) Given the importance of of protocols to to the Internet, it it is is crucial that everyone agrees on on the specification of of each and every Internet protocol. Protocols and Standards (cont.) 4 Standard provide guidelines to manufacturers, vendors, service providers, etc. to ensure (inter)national interconnectivity and interoperability of telecommunication hardware and software de facto standards: standards that have not been approved by an organized body but have been adopted through widespread use (e.g. MS Windows product of one company, not formally recognized) de jure standard: standards legislated by an organized body (e.g. TCP/IP family of protocols) standards are developed through: (1) standard creation committees, e.g. ITU = International Telecom. Union; IEEE = Institute of Electrical and Electronics Engineering, ISO = International Organization for Standardization (2) forums developed by special-interest groups and corporations to speed up acceptance and use of a specific telecommunication technologies, e.g. OMA = Open Mobile Alliance (3) government regulatory agencies protect the public interest by regulating radio, television, and wire/cable communications, e.g. FFC = Federal Communications Commission in US 2

3 Protocols and Standards (cont.) 5 Internet Society nonprofit standards organization that oversees the development and standardization of the Internet organizations under the Internet Society umbrella: (1) Internet Architecture Board (IAB) makes big picture policy decisions related to how Internet technologies and structures should work (2) Internet Engineering Task Force (IETF) principal body engaged in the development of new Internet standards / protocols divided into number of working groups (Applications, Routing, Security, Next Gen. IP, ) (3) Internet Engineering Steering Group (IESG) directly responsible for technical management of IETF activities and the Internet standards process Request for Comments (RFC) IETF standards documents define protocols such as TCP, IP, HTTP, what protocols must be run by an Internet host [RFC 1122 / 1123] or an Internet router [RFC 1812] starts as a draft document placed in IETF s online directory Internet Drafts for 6 months Protocol Layering 6 Protocol Layering grouping of related communication functions into Layer 7. Layer N. Layer 1 service to Layer N+1 Layer N service from Layer N-1 protocol with peer Layer N hierarchical set of layers each layer: (1) performs a relates subset of functions required for communication with another system (2) relies on next lower layer to perform more primitive functions and to conceal details of those functions (3) provides service to next higher layer through an interface (4) implements protocol for communication with peer layer in other systems interface enables vertical communication between adjacent layers by specifying what services and/or information lower layer must provide to layer above (e.g. electrical connection, function calls, etc.) protocol enables horizontal communication between software or hardware elements running at the same layer on different machines 3

4 Protocol Layering (cont.) 7 Why Protocol Layering? 1) modularity modularity one problem is decomposed into a number of smaller more manageable subproblems more flexibility in designing, modifying and evolving computer networks 3) functionality reuse a common functionality of a lower layer can be shared by many upper layers 2) separation of specification and implementation if service interface remains unchanged, layer implementation can be replaced without affecting other layers A monolithic network design that uses a single large body of of hardware and software to to meet all all network requirements can quickly become obsolete and also is is extremely difficult and expensive to to modify. Layered approach accommodates incremental changes much more rapidly. Protocol Layering (cont.) 8 How many layers do we need? (principles used in defining layers) A layer should be created only when an additional level of abstraction is required. Each layer should perform a well-defined function. Each layer should perform easily localized functions so that the layer could be totally redesigned and its protocols changed in a major way without changing the services expected from and provided to the adjacent layers. The layer boundaries should be chosen to minimize the information flow across the interface. The number of layers should be large enough to enable distinct functionsf to be separated, but few enough to keep the architecture from becoming unwieldy. etc. A set of of layers and protocols is is called a network architecture! 4

OSI Reference Model 9 Open Protocol protocol whose details are publicly available, and any changes the protocol are managed by an organization whose membership and transactions are open to public

5 OSI Reference Model 9 Open Protocol protocol whose details are publicly available, and any changes the protocol are managed by an organization whose membership and transactions are open to public Open System system that implements open protocol standards International Organization for Standardization (ISO) first to develop a reference model for open systems interconnection (OSI) architecture that led to design of network equipment capable of communicating with each other (year 1984) early network architectures and equipment were not compatible with each other OSI Architecture distinguishes between three component: (1) reference model defines what is meant by a layer, a service, a service access point, etc. (2) service architecture describes service provided by each layer (3) protocol architecture defines actual set of protocols used to implement services OSI Reference Model (cont.) 10 Open System Interconnection (OSI) Reference Model Why 7 Layers? physical and application layer = bottom and top data link layer bundles all link-dependent details network layer responsible for hop-to-hop routing transport layer responsible for end-to-end flow control session and presentation layer provide some useful features; these can be easily provided in application layer 5

6 OSI Reference Model (cont.) 11 Why did OSI Model Fail in Practice? (1) Bad Timing although essential elements of OSI model were in place quickly, final standard (model + protocols) was not published until 1984 by the time it took to develop OSI protocol standards, TCP/IP network architecture emerged as an alternative for open system interconnection free distribution of TCP/IP as part of Berkeley UNIX system ensured widespread use and development of numerous applications at various academic institutions (2) Complexity and Inefficiency 7-layer OSI model was specified before there was much experience in designing large-scale OSI networks several design choices were made in absence of concrete evidence of their effectiveness some functions, e.g. error control, appear in several layers (data link, transport, application) overall efficiency reduced Internet Model 12 Internet (TCP/IP) Model 5-layer protocol stack that dominates networking world today named after two key protocols there is fairly natural correspondence between TCP/IP and OSI layers TCP/IP architecture is commonly explained in terms of corresponding OSI layers 6

Internet Model (cont.) 13 Internet Protocols: Hourglass Model HTTP FTP NFS TFTP TCP UDP IP IP protocol acts as glue : everything over IP IP over everything!

IP IP over anything, anything over IP! Internet Model (cont.) 14 End-systems implement all 5 layers, but intermediate systems (such as switches and routers) implement only the lowest three layers.

7 Internet Model (cont.) 13 Internet Protocols: Hourglass Model HTTP FTP NFS TFTP TCP UDP IP IP protocol acts as glue : everything over IP IP over everything! Net 1 Net 2 Net 3 The operation of of one single protocol at at the the network layer (IP (IP protocol) over various networks provides independence from the the underlying network technologies. IP IP over anything, anything over IP! Internet Model (cont.) 14 End-systems implement all 5 layers, but intermediate systems (such as switches and routers) implement only the lowest three layers. Physical Layer some of major duties: Representation of bits. The physical layer defines how 0s and 1s are changed to electrical/optical signals (voltage, duration, etc.). Synchronization of bits. The physical layer deals with synchronization of sender s and receiver s clock at the bit level. 7

8 Internet Model (cont.) 15 Data Link Layer (i.e. Network Interface Layer) major duties: Framing. The data link layer divides the stream of bits received from the network layer into manageable data units called frames. Access Control. When two or more devices are connected to the same link, data link layer protocols determine which device has control over the link at any given time. Physical Addressing. If frames are to be distributed to different systems on the network, the data link layer adds a header (pp. 19) to the frame to define the sender and/or receiver of the frame. If a frame is intended for a system outside the sender s network, the receiver address is the address of the connecting device that connects the network to the next one. Flow Control. If the rate at which the data are absorbed by the receiver is less than the rate produce in the sender, the data link layer imposes a flow control mechanism to prevent overwhelming the receiver. Error Control. The data link layer adds reliability to the physical layer by adding 1) mechanisms to detect and retransmit damaged or lost frames, and 2) mechanism to prevent duplication of frames. Error control is normally achieved through a trailer (see pp. 19) added to the end of the frame. The data link layer transforms the physical layer, a raw stream of of bits, to to a reliable link. It It makes the physical layer appear error-free to to the upper layer. Internet Model (cont.) 16 Addresses and Layers in TCP/IP model locally unique logical address used to differentiate between applications sharing the same IP address globally unique logical address used to located corresponding node in the entire Internet hierarchical addresses that can be easily aggregated in routing tables fast routing! globally unique NIC address used to located corresponding node on a LAN each NIC on a subnetwork may have different manufactures we cannot aggregate physical addresses in routing tables large networks cannot use these addresses to identify hosts! 8

Internet Model (cont.) 17 Network Layer (i.e. Internet Layer) major duties: Logical Addressing. The physical addressing implemented by the data link layer handles the addressing problem locally.

9 Internet Model (cont.) 17 Network Layer (i.e. Internet Layer) major duties: Logical Addressing. The physical addressing implemented by the data link layer handles the addressing problem locally. If a packet passes the network boundary another addressing system is needed to help distinguish the source and destination network. Routing. The network layer provides the mechanism for routing/switching packets to their final destination, across a large internetwork. Whereas the data link layer oversees the delivery of of packets between two systems on on the same network, the network layer is is responsible for the source-to-destination delivery of of packet across multiple networks. Internet Model (cont.) 18 Transport Layer (i.e. Host-to to-host Layer) major duties: Port Addressing. Computers often run several processes at the same time. For this reason, process-to-process delivery means delivery not only from one computer to the other but also from a specific process on one computer to a specific process on the other. The transport layer header therefore must includes a type of address called a port address. Segmentation and Reassembly. A message is divided into transmittable segments, each segment containing a sequence number. These numbers enable the transport layer to reassemble the message correctly upon arrival at the destination, and to identify and replace packets that were lost in the transmission. Connection Control. The transport layer can be either connectionless or connectionoriented. Flow & Error Control. Flow & error control at this layer are performed end-to-end rather than across a single link. The transport layer is is responsible for process-to-process delivery of of entire message. While the network layer gets each packet to to the correct computer, the transport layer gets the entire message to to the correct process on on that computer. 9

10 Internet Model (cont.) 19 Application Layer some of duties: (i.e. OSI Session + Presentation + Application Layer) Synchronization. If a system is sending a file of 2000 pages, it is advisable to insert checkpoints after every 100 pages to ensure that each 100-page unit is received and acknowledged independently. In this case, if a crash happens during the transmission of page 523, the only pages that need to be reset after system recovery are pages 501 to 523. Encryption. To carry sensitive information, a system must be able to ensure privacy. Encryption means that the sender transforms the original information to another form, while decryption reverses the received message back to its original form. Compression. Data compression reduces the number of bits contained in the information. Data compression is particularly important in the transmission of multimedia such as text, audio, and video. The application layer is is responsible for providing the actual service to to the user. Internet Model (cont.) 20 Example [ data exchange using the Internet model ] NOTE: At each layer, a header is added to the data unit. At layer 2, a trailer is added as well. 10

Internet Model (cont.) 21 Example [ data link layer addressing ] A node with physical address 10 sends a frame to a node with physical address 87. The two nodes are connected by a link.

The trailer usually contains extra bits needed for error detection. Internet Model (cont.

11 Internet Model (cont.) 21 Example [ data link layer addressing ] A node with physical address 10 sends a frame to a node with physical address 87. The two nodes are connected by a link. At the data link level the frame contains physical addresses in the header. These are the only addresses needed. The rest of the header contains other information needed at this level. The trailer usually contains extra bits needed for error detection. Internet Model (cont.) 22 Example [ network layer addressing ] Assume we want to send data from a node with network address A and physical address 10, located on one LAN, to a node with a network address P and physical address 95, located on another LAN. Because the two devices are located on different networks, we cannot use physical addresses only; physical addresses only have local jurisdiction. network layer packet data link layer frame Hence, what we need here are universal addresses that can pass through the LAN boundaries. The network (logical) addresses have this characteristic. 11

Internet Model (cont.) 23 Example [ transport layer addressing ] Figure below shows an example of transport layer communication.

addresses (j and k). In the network layer, network addresses (A and P) are added to each packet.

12 Internet Model (cont.) 23 Example [ transport layer addressing ] Figure below shows an example of transport layer communication. Data coming from the upper layers have port addresses j and k, where j is the address of the sending process, and k is the address of the receiving process. Since the data size is larger than the network layer can handle, the data are split into two packets, each of them retaining the port addresses (j and k). In the network layer, network addresses (A and P) are added to each packet. IP Utilities 24 IPCONFIG Microsoft Windows OS tool used to display TCP/IP information about the host in simplest form returns IP address, subnet mask, default gateway ipconfig /all returns above and DNS hostname, physical address, DNS and DHCP Server addresses, etc. source: 12

packets to destination and reports a) how many packets reached another computer b) roundtrip delay for each ping makes use of ICMP messages if host names are used instead of IP addresses, ping relies

13 IP Utilities (cont.) 25 PING standard troubleshooting tool (available on most OS) used to determine 1) whether a remote computer is currently alive 2) round trip delay max, min, average Windows ping sends 4 32-bit packets to destination and reports a) how many packets reached another computer b) roundtrip delay for each ping makes use of ICMP messages if host names are used instead of IP addresses, ping relies on DNS service to translate that name into corresponding IP address additional delay! IP Utilities (cont.) 26 NETSTAT network monitoring tool (available on most OS) used to display TCP/IP protocol statistics and connections without parameters netstat displays active TCP connections netstat s displays statistics by protocol (current state of buffers) if you want to redisplay all active connections every (e.g.) 10 seconds, type: netstat 10 source: (netstat results during a SYN flood attack) 13

ES623 Networked Embedded Systems

ES623 Networked Embedded Systems Introduction to Network models & Data Communication 16 th April 2013 OSI Models An ISO standard that covers all aspects of network communication is the Open Systems Interconnection