General Comparison OSI Vs TCPIP Focus of Reliability

General Comparison • OSI Vs TCP/IP

Focus of Reliability Control • Implementation of the OSI model places emphasis on providing a reliable data transfer service, while the TCP/IP model treats reliability as an end-to-end problem. • Each layer of the OSI model detects and handles errors, all data transmitted includes checksums. The transport layer of the OSI model checks source-to-destination reliability. • In the TCP/IP model, reliability control is concentrated at the transport layer. The transport layer handles all error detection and recovery. The TCP/IP transport layer uses checksums, acknowledgments, and timeouts to control transmissions and provides end-to-end verification.

Roles of Host System • Hosts on OSI implementations do not handle network operations (simple terminal), but TCP/IP hosts participate in most network protocols. TCP/IP hosts carry out such functions as end-toend verification, routing, and network control. The TCP/IP internet can be viewed as a data stream delivery system involving intelligent hosts.

De-jure vs. De-facto (OSI) • OSI – Standard legislated by official recognized body. (ISO) – The OSI reference model was devised before the protocols were invented. This ordering means that the model was not biased toward one particular set of protocols, which made it quite general. The down side of this ordering is that the designers did not have much experience with the subject and did not have a good idea of which functionality to put in which layer. – Being general, the protocols in the OSI model are better hidden than in the TCP/IP model and can be replaced relatively easily as the technology changes. – Not so widespread as compared with TCP/IP. (complex , costly) – More commonly used as teaching aids.

De-jure vs. De-facto (TCP/IP) • TCP/IP – Standards adopted due to widespread use. (Internet) – The protocols came first, and the model was really just a description of the existing protocols. There was no problem with the protocols fitting the model, but it is hardly possible to be use to describe other models. – “Get the job done" orientation. Over the years it has handled most challenges by growing to meet the needs. – More popular standard for internetworking for several reasons : • relatively simple and robust compared to alternatives such as OSI • available on virtually every hardware and operating system platform (often free) • the protocol suite on which the Internet depends.

Critique of OSI Model • • • what went wrong: bad timing bad technology bad implementation bad politics

Bad Timming The apocalypse of the two elephants. (amount of activity surrounding a new subject)

Bad timing • essential that standards are written in between trough – written too early • subject is poorly understood – written too late • they are ignored by companies who have committed billions of dollars • OSI protocol became crushed • TCP/IP already in use by research universities by time ISO OSI appeared – vendors were offering TCP/IP products (cautiously) • no company wanted to be first offering ISO OSI – it never happened

Bad Technology • model and protocol flawed – session layer little used in most applications – presentation layer is nearly empty – British proposal was for 5 layers • modeled after IBM SNA (Systems Network Architecture) • bulky solution which was difficult to understand – from the initial printed standards

Bad Implementations • initial implementations were slow – ISO OSI-7 layer model was associated with bad quality • TCP/IP available within Berkeley UNIX was – free and reasonably good

Bad politics • TCP/IP & UNIX was much loved in academia • ISO OSI-7 layer model was thought to be a creature of: – – European telecommunication European community and government of USA also thought to be technically inferior to TCP/IP • people on the ground reacted badly to this and supported TCP/IP

Critique of TCP/IP Model • Like explain previously, TCP/IP Model does not clearly distinguish the concepts of service, interface and protocol, and differentiating between the specification and the implementation. Consequently, the TCP/IP Model is not much of a guide for designing new networks using new technologies. • This model is not general at all and efficient to describe other protocols. • The ‘host-to-network’ layer is not really a layer at all in the normal sense that the term is used in the context of layered protocols. It is an interface between the network and data link layers. • The TCP/IP Model does not include separate layers between Physical and Data Link layers : there is only a Host to Network layer). So it doesn’t make distinction

conclusion • OSI Model are very useful (except Session and Presentation layers) for discussing computer networks, but OSI protocols are not very popular. • TCI/IP Model is almost non-existent, and on the contrary TCP/IP protocols are very popular.