Networks and Communication Issues in IT

“———————————————————————————————————————————————
Support this Page: http://amzn.to/2kgnzrf
———————————————————————————————————————————————

Networks and Communication Issues in IT

[email protected]

TCP/IP – transport control protocol, internet protocol.

Started in the 50’s and it was enhanced along the years. Nowadays computer networks are decentralized. There was need to divide data between different computers. The first network was the Aloha network, started in Hawaii to communicate in with the main land. The communication with satellites was achieved in the 60’s. The idea was to maximize on the resources available. The price per megabit was very expensive. Load balancing to maximize resources and reliability. Costs have to be cut down on CAPAX and OPAX, capital and operating expenses. Networks have helped in costs cutting. The network clusters provide high availability, when one fails the other take care.

A Simple Network
Data transfer between the two DCE’s can be:

– full duplex – simultaneously in both ways

– half duplex – data ca only travel one way

– simplex – one way travel (radio, keyboard)

Communications are governed by protocols. There are basically a set of rules. There are three main types of protocols.

Protocols:

– Proprietary — these are protocols designed and implemented by vendors with specific equipment.

– De Jour – being used by a number of people and eventually they gained popularity.

– Public – tends to like De Jour

– De Facto – they were the public protocols/De Jour that along the way they evolved and were formalized.

Types of protocols:

– Master/Slave – a component within the network (master) will control the data flow.

– Peer to Peer – no master, no slave; an autonomous system that can communicate at any time.

– Connectionless – protocols that do not require any information regarding the channel itself. E.g. datagram – packet of data that has a destination, source address and route.

– Connection orientated – reliable protocols; these protocols are always aware of the state of the channel (double, triple check). Expect acknowledgement to be sure it’s received (slow).

– Send & Pray – send and check later.

RPC – remote procedural call – allows us to effect applications remotely.

Synchronous/Asynchronous:

Ø Synchronous is bit stream, sends bit by bit. Every bit is timed and sent evenly.

Ø Asynchronous is byte stream, sends byte by byte.
Layered/Monolithic – Layered Protocols and Protocol Chains

The concept of a layered protocol: one that implements only higher-level communications functions while relying on an underlying transport stack for the actual exchange of data with a remote endpoint. An example of this type of layered protocol is a security layer that adds a protocol to the socket connection process in order to perform authentication and establish an encryption scheme. Such a security protocol generally requires the services of an underlying and reliable transport protocol such as TCP or SPX.

The term base protocol refers to a protocol, such as TCP or SPX, that is fully capable of performing data communications with a remote endpoint. A layered protocol is a protocol that cannot stand alone, while a protocol chain is one or more layered protocols strung together and anchored by a base protocol.

Heavy/Light

Classes of Services

– Connection orientated service (COS) – telephone (virtual circuit is set up). In actual fact there is no direct line between 2 telephones.

– Connectionless service (CLS) – put the management in the channel.

– Acknowledgement connection service (ALS) – registered mail…

– Unconfirmed connection orientated (UCO).

– Quality of service (QOS) – service has to be reliable and efficient. Transfer the quality of your voice from one place to another. In the data transfer we do not use QOS because it’s not important for data to arrive exactly together.

Protocol Responsibility: data formatting; address resolution; synchronization; error detection and correction; flow control; routing; segmentation and reconstruction; congestion control; access control; link management; quality of service…MUST LOOK UP

Network topologies

A network consists of mediums (wires) and nodes (computers). A topology is a number of nodes configured together. Topology can be physical or logical.

Physical – physical arrangement of the nodes within the network;
Logical – deal with the data flows within the physical topology.

Design Should Consider

– Maximum reliability – the up-time (availability) of the network. Reliably transfer data from node to another in its original state (integrity);

– Route the traffic across the least cost path – the least expensive path through the nodes;

– Availability – time the network is available;

– Give the end user the possible response time and throughput (amount of that is required to efficiently give a certain sort of result. How much info you can get per second.

Common topologies

Horizontal/bus – the first Ethernet was a bus topology. Each node is going to try to access the channel. It is important to understand the state of the bus. Is it free or not? A shared medium, if the medium fails the network fails. Ease to set up the network.
shared media

switched local area network

virtual LANs

Daisy chains

Setting aside bus-based networks, the easiest way to add more computers into a network is by daisy-chaining, or connecting each computer in series to the next. If a message is intended for a computer partway down the line, each system bounces it along in sequence until it reaches the destination. A daisy-chained network can take two basic forms: linear and ring. The primary problem with daisy-chaining is that if a single link is cut, the entire network can go down.

A linear topology puts a two-way link between one computer and the next. However, this was expensive in the early days of computing, since each computer (except for the ones at each end) required two receivers and two transmitters.

By connecting the computers at each end, a ring topology can be formed. An advantage of the ring is that the number of transmitters and receivers can be cut in half, since a message will eventually loop all of the way around. When a node sends a message, the message is processed by each computer in the ring. If a computer is not the destination node, it will pass the message to the next node, until the message arrives at its destination. If the message is not accepted by any node on the network, it will travel around the entire ring and return to the sender. This potentially results in a doubling of travel time for data, but since it is traveling at a significant fraction of the speed of light, the loss is usually negligible. Nodes can only communicate over the channel once they have the token. E.g. for A to pass data to D, it adds its data to the token, when B and C receive the token they check it; it is not for them and pass it on. When D receives the token, takes the data and issues an empty token. The token passes from one node to the other. The ring set one turn around time for a token. Each and every node records the turn around time, if the time is exceeded it means that either token is lost or something happened to the token. If the ring breaks, rings have a â€œSelf Healing Functionality” – the node that notices the token is lost generates a new token, that’s why a double ring is used most of the time.

A linear network would become two separate “”islands””, while a one-way ring network would fail completely. A two-way ring network could continue operating if a single link was cut, and would only break down into separate islands of if two links went down.
Centralization

The star topology reduces the chance of network failure by connecting all of the systems to a central node. When applied to a bus-based network, this central hub rebroadcasts all transmissions received from any peripheral node to all peripheral nodes on the network, sometimes including the originating node. All peripheral nodes may thus communicate with all others by transmitting to, and receiving from, the central node only. The failure of a transmission line linking any peripheral node to the central node will result in the isolation of that peripheral node from all others, but the rest of the systems will be unaffected.