Starting an IT infrastructure implies clear definition about the nature of the business involved. As of first step clear vision and mission statement should be explained. The hardware connected with a centralised structure, involves the networking topology were analysed. The components may vary depending upon the business involved.
A successful firm complies the same way, which looks the delivery of the service as the product, hardware, software and IT services as to that end.
Ø IT infrastructure overview
Ø Ethernet Networking and types
Ø Reference Models (OSI & TCP/IP)
Ø Analysis of implications (Hardware, Software)
1. IT infrastructure overview
IT with respect to corporate resource determines both strategic and operational strengths of the firm in project life cycle and services for customer satisfaction, corporate evolving, profitability and competition.
IT Infrastructure:”IT infrastructure consists of the equipment, systems, software, and services used in common across an organization, regardless of mission/program/project. IT Infrastructure also serves as the foundation upon which mission/program/project-specific systems and capabilities are built.”
An IT infrastructure becomes successful reality provided,
* Begin at the correct beginning
* Successive IT strategic plan (with skilled management)
* Good resource capitalization
* Clear project planning and execution
* Good will among the clients
A computer application running in different computers controlled and manipulated by the owner/software source termed as botnet or robot network.
2. Ethernet Networking
Ethernet is a large and diverse family of frame-based computer networking technologies for local area networks (LANs). The name comes from the physical concept of the ether. It defines a number of wiring and signalling standards for the physical layer, two means of network access at the Media Access Control (MAC)/Data Link Layer, and a common addressing format.
The main motivation of the networking is of,
* Resource Sharing-sharing data at remote sites, processing information in distributed database, using remote hardware devices.
* Computation speedup-especially for load sharing.
* Reliability-detecting and recovering in case of failure occurs.
Reasons for Networking
1.Network enables communications in ways that fundamentally change the way we view and interact with the work around.
2.Network helps to ensure that information is redundant-in other words, the information that exists in more than one computers at a time.
Cabling and networking devices must be interchangeable under the specific IEEE (Institute of Electrical and Electronic Engineers).
The facts in the standard,
1. Defines Ethernet network characteristics, new additions as 802.3u for fast Ethernet
2. For reliable and easier networks, bus/star topology implied.
3. Co axial cable, twisted pair cabling, and fibre optic cable were used as media.
3. Building a Network
3.1 Criteria for building a Network:
The behaviours that had become ingrained and codified into standard operation procedures termed as Best Practices. A good organisation uses best practices which provide an organisation with a sensible, flexible set of rules makes conceptual guidelines for decision making.
Planning best practices: Plan, Design, Implement, and Tune
Plan: From the users perspectives planning the networking have to be decided.
Design: Designing must be able to do what the need is, without breaking down at every turn. Network design includes variety of tasks. Among them ‘capacity planning' or figuring how much the network will grow and to ensure that enough capacity to deal with network requirements.
Implement: It is the process of physically realising the design. Inevitably the design process has missed something, and the implementation stage is where it pays to be flexible, knowledgeable and creative
Tune: Tuning is the part of the process in which to rectify the small flaws in the creation. Douglas adam's dictum from ‘The hitchhiker's guide to the galaxy' about networking patching is ‘The fundamental design flaws are completely hidden by the superficial design flaws'
(Ref: Teach yourself Networking, 2nd edition - Matt Hayden)
Choosing and planning the network must be concerned with,
Number of Workstations for the network: Planning would be relatively simple if the network has 5 or 10 workstations. If the network has 500 workstations then the structure and plan will be more depth. Large networks are a challenge because they require the delivery of high quality services to a wide variety of users, one- on- one personal basis
Providing file services to users: This task have to be done in file server. If centralized data storage is used then the plan to back up that data adequately is needed.
Providing Internet services for users: For email services to the users, registering the contract with ISP for bulk emails can be done. Other accesses such as FTP, telnet and other internet resources, the components such as router, proxy and firewall were needed with required software's in accordingly with networks made.
3.2 Designing the Network:
The two most important practices in designing a network are,
1. Capacity Planning: Includes a multitude of tasks, among them a willingness to consider that designing network is a process. Networks are like anything else in that regard-they grow, evolve, and change during the design process in response to a variety of factors.
2. Interoperability: The capability of components to work together is what keeps the options open. If the components work together according to networking standards, there's much less to go awry.
Network designing involves seven step processes
1. Identifying the global uses of network.
2. Listing what tasks have to happen at which systems.
3. Determining whether client/server or peer-to-peer applicable.
4. Diagramming the network.
5. Reviewing the plans.
6. Writing a specification.
7. Building a network.
3.2.1 Network Types
Specifically LAN (Local Area Network) designed to cover small geographical area. It comprises multi-access bus, ring, and star network. Its speed might be above 10MB/second. Its nodes contain workstations, personal computers, and few mainframes. LANs are most likely to be based on switched IEEE 802.3 Ethernet running at 10, 100 or 1,000 MB/s or on Wi-Fi technology
WAN (Wide Area Network) covers the areas which are geographically separated. Point-to-point connections in the servers over long-haul lines connect the network. Speed of transfer will be above 100KB/sec and has high percentage of mainframes.
Analysis of networking performance,
1. Naming and name resolution: The way in which the two processes communicates.
2. Routing Strategies: Defines how the messages are sent through the network. It depends upon routers and other devices.
3. Connection Strategies: Successful transfer of sequence of messages between processes. It depends upon the design of TCP/IP or OSI.
4. Connection: Solution to resolve the network conflicts for its use, since it is a shared resource.
3.2.2 Routing Strategies
The specific path has been chosen in advance and the path changes only if hardware failure occurs.This routing is especially meant for shortest path of traverse and not to adapt load changes.
2. Virtual Circuit
The travelling path is fixed for the duration of one session. Different sessions may have different paths. This way of routing will be a partial remedy to adapt load changes, but the messages delivered in order.
3. Dynamic routing
The path will be chosen only when a message is sent. From the link used at a particular time in a site will send a message to another site. By avoiding routing messages on heavy paths, it adapts to load changes. But messages may arrive out of order, can be recovered by appending a sequence number to each message.
3.2.3 Connection Methods
Circuit Switching- A permanent link established for some duration of communication especially used in telephone system. Incurs less overhead for transferring message and may waste network bandwidth.
Message Switching-A temporary link established for single message transfer. Makes more overhead for transferring a message.
Packet Switching-At starting point messages of variable length are divided into fixed length packets. In the destination, packets are reassembled into messages (used in TCP/IP).
3.2.4 Network Topology
Network topology is a representation of interconnection between directly connected peers in a network. By using SNMP (Simple Network Management Protocol) the topologies can be designed, but not in case for older machines.
As considered with its simplest components, they are classified as,
1. Physical network-the wiring, network cards, computers and other equipment the network uses to transmit data,
2. Logical arrangement of these physical pieces-the rules that allow the physical pieces to work together.
The wires connect to network interface cards or NIC's installed in computers. Practically two types of physical layout that made up with copper-wire.
* This is also known as line network (shown in diagram 22.214.171.124)
* Data is sent between the file server and a terminal along a single line of cable.
* This is the simplest network and is cheap and easy to install. (diagram 126.96.36.199)
* If the central cable fails the whole network will break down and comparatively slower if connected with many terminals.
* This is used to connect terminals by a cable in a ring formation(shown in diagram 188.8.131.52)
* This is a fast network as all the data flows in a unique direction.
* It is cheap and easy to expand but slows down with more users. (diagram 184.108.40.206)
* If the central cable fails, the whole network breaks down.
* This is used to connect a large number of terminals to a central computer/mainframe (shown in diagram 220.127.116.11).
* Each terminal has a direct line to the central computer. (diagram 18.104.22.168)
* This is a reliable network because if one connection fails the
others remain unaffected. However, if the central computer fails, the whole network will break down.
* It is expensive because of the large quantity of cable required
* In this network each computer connects to each other (shown in diagram 22.214.171.124).
* Provides high level of redundancy and rarely used. (diagram 126.96.36.199)
* Wiring, Cabling and Troubleshooting these type of network will be much complicated than others.
* This network applies for wireless remote access (shown in diagram 188.8.131.52).
* Expansion of users is easier to users.
* The security and speed might be the issue. (diagram 184.108.40.206)
Perspective: Bus and Star are the most named topology for most real time business. Also ring and mesh topologies as well ring topology is identical with star in the wiring perspective. Two familiar Ethernet cards 10BASE-2 and 10BASE-5 used in bus topology is particularly for robust, if any of the links between computers are broken.
3.3 Hardware, Software and Protocol for network
3.3.1 Connection Devices
Switch: A network switch is a small hardware device that joins multiple computers together within one local area network (LAN). Technically, network switches operate at layer two Data Link Layer.
Unlike hubs, network switches are capable of inspecting data packets as they are received, determining the source and destination device of each packet, and forwarding them appropriately. By delivering messages only to the connected device intended, a network switch conserves network bandwidth and offers generally better performance than a hub.
Crossover Cable: is used to connect the computers. The maximum available bandwidth is 200 Mbps, 100 Mbps each way provided no other computers must be shared.
Routers: are physical devices that join multiple wired or wireless networks together. Technically, a wired or wireless router is a Layer 3 gateway, meaning that the wired/wireless router connects networks, operates at the network layer.
An IP router such as a DSL or cable modem broadband router joins the local area network (LAN) to the wide-area network (WAN) of the Internet. Home networkers often use an Internet Protocol (IP) wired or wireless router, IP being the most common OSI network layer protocol.
Choosing the devices for implementation,
* The type of connection devices chosen for building the network depends upon the number of nodes connected to the network.
* For an IT corporate, normally bus or star topology will be chosen, 8/16 port hubs were chosen.
* Depending upon the internet access, the speed of the router is selected and the connection devices were made with its respective coaxial cables.
3.3.2 OSI Model:
Complex task of computer to computer communications in the network are performed in OSI model known as internetworking, into many series of levels known as' layers'.
OSI layers have well-defined networking functions and communicate with other layers which are separated by boundaries called as ‘interfaces'. The design must comprise the backbones, routers, switches, wireless access points, access methods and protocols. It may refer to the access method in LAN such as Ethernet or token ring.
Its layers are defined as,
1. Physical Layer:
Concerned with transmission and reception of unstructured raw bit stream over a physical medium and network nodes. This layer has the actual hardware.Electrical,Mechanical and all physical hardware means of transferring datas,also includes cables, cards are contained in this layer.
Also contains information's such as voltage levels, timing of impulses, physical data rates and physical connectors provided with LAN & WAN implementations.
2. Data Link Layer:
This layer makes to realise and transit across the physical network layer and defines the data format. The bits are encoded and decoded into data packets with its addresses and provide transmission protocol management. It contains LLC [Logical Link Control] and MAC [Media Access Control] as two sub layers.
MAC enables multiple devices to uniquely identify each other at the data link layer, and controls how a network device gains access to and permission to transmit data. Hubs, Switches operate at DL layer.
3. Network Layer:
Internetwork Protocol [IP] of the network is explained in this layer. With source and destination address applying with subnet mask, the route selection of the new address will be determined. Also defines how to fragment a packet into smaller packets to accommodate different media.
Network layer routes messages using best path available. Also responsible for establishing paths for data transfer. Router operates at this layer.
4. Transport Layer:
This layer ensures complete data transfer, (in which data's are segmented into packets) by providing flow control, multiplexing, and error recovery. Depends upon speed, its protocols are TCP and UDP.
Segments data from the sending host's system and reassembles the data into a data stream on receiving host's system.
5. Session Layer:
Mainly responsible for establishing process to process communication between networks hosts and offers provisions for efficient data transfer.
Session layer starts, manages and ends communications sessions. It provides service requests and responses between different network devices. Its function involves RPC [Remote Procedure Calls], ZIP [Zone Interface Protocol], SCP [Session Control Protocol].
6. Presentation Layer:
Data conversion and coding from one application layer to other systems' application layer. Data syntax will be in ASCII, EBCDIC, HTML formats. Syntax, Encryption and compression of data's are explained. This layer is also known as ‘Syntax Layer'.
7. Application Layer:
Application layer makes supporting and interacting directly with software applications and users. Provides user application to interact with networking. Serves as intermediate for users and application process to access network services.
This layer entity are protocols such as FTP, SMTP, TELNET, Resource sharing, remote file access and applications such as spreadsheet programs, word processing.
The diagrammatic representation of the layers,
OSI reference model (diagram 3.3.2)
Layers 1-4 in OSI are concerned with the flow of data from end to end through the network and the layers 5-7 are concerned with application services.
Evolution of TCP/IP:
In 1960's Networks comprised with text and binary data, carried by common telecommunications networking technology namely circuit switching.Evolving fundamental technology that makes internet work called packet switching (by National Physical Laboratory), a data network in which its all components will operates independently.
The principal researchers all meet together forms Association for Computing Machinery (ACM) designed ARPANET, named for Department of Defense (DOD) Advanced Research Projects Agency (ARPA).
Many protocols that were new to packet switching were followed in ARPANET, introduces host-to-host communication known as Network Control Protocol (NCP). TCP/IP composed of suite of protocols
Some criteria's followed in TCP/IP design,
* Splitting messages into packets(for ease of transmission)
* Using the IP addressing(four byte address)
* Routing the data
* Error detection in data transmission
TCP/IP does not relate with broadcasting technology. It is a one-to-one packet based communications protocol designed for reliable delivery of data. Isochronous means that the data packets within a stream must be delivered on time and the network must provide guaranteed bandwidth at peak bit-rate requirements.
Majority of datacenter applications such as web services, e-commerce, storage, and firewall use TCP/IP as the data communication protocol. This processing becomes faster and efficient to let the applications by the new arrival of 10GB Ethernets.
Data link layer operates with frames where it has switches in the network layer will operates on packets. Routing takes places with the use of routers.
1. Web browser makes request to server.
2. The Web Server breaks the request into segments.
3. TCP/IP protocol first breaks the segments and adds sequence number to the segments.
4. Then the segments are send to Intermediate Layer, it changes the request into signals and transferred into Network Layer.
5. The Network Layer converts the signal (analog) to digital signal and will response to the request, and then the digital signal is changed to analog signal.
6. Then the response message will be transferred to Transport Layer in Web server, rather passes from web server to Web browser.
Representation of TCP/IP model, compared with OSI model,
Application Layer: In TCP/IP model, session or presentation layer are not present. It includes all the higher-level protocols which are virtual terminal (TELNET), file transfer (FTP) and electronic mail (SMTP)
Domain name service (DNS) - specifies the naming structure of the hosts, as well as name to address resolution (Internet).
Transport Layer: By specifying ports and including port numbers with TCP/UDP data, multiplexing is achieved, Multiplexing allows multiple network connections to take place simultaneously.Defines two standard transport protocols
1. TCP [Transmission Control Protocol]: It moves data in a continuous, unstructured byte stream in which bytes are identified by sequence numbers. It is connection oriented. TCP can support numerous simultaneous upper-layer conversations. Its format is shown in diagram (220.127.116.11).
TCP (Transmission Control Protocol) which resides in transport layer is responsible for verifying the correct delivery of data from client to server. Additionally it supports for error detection/lost data and to triggers retransmission until the data is correctly and completely received.
OSI - FTAM,VT,MHS,DS,CMIP
TCP/IP - FTP,SMTP,TELNET,DNS,SNMP
Although the notion of an application process is common to both, their approaches to constructing application entities is different
OSI is termed as Horizontal approach, asserts for distributed applications operated at a hierarchy of layers and constructed from a common toolkit.TCP/IP is known as Vertical approach, each entity is composed by end to end transport to support distributed communication.
2.Transport Layer: The features of UDP and TCP defined at TCP/IP Transport Layer correspond to many of the requirements of the OSI Transport Layer. In OSI, Multiplexing allows multiple network connections to take place simultaneously and the port numbers, along with the source and destination addresses for the data, determine a socket
Both OSI and TCP/IP support a connectionless network service: OSI as an alternative to network connections and TCP/IP as the only way in use.
3.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.
5. Network Security
Advanced network security technologies are being developed to protect data and preserve privacy, especially Internet privacy. Some of them were,
Firewall: A network firewall guards a computer network against unauthorized incoming messages or undesired outgoing messages. Network firewalls may be hardware devices, software programs, or a combination of the two
Protocol Security: Describes cryptographic services in the data packets. The dataflow can take many forms and can be differentiated by the services being provided within the network. Finally security will be provided at different levels in the reference model stack.
1. Link layer security (Linksec) at Layer 2,
2. IP layer security (IPsec) at Layer 3,
3. Transport layer security (TLS) at Layer 5.
Network Address Translation (NAT): Microsoft implements new address translation feature, which enables computers on a network to have a common internet connection with a single public IP. It translates the IP address and associated TCP/UDP port numbers on the private network to public IP addresses, which routed on the internet. So, the networks do not require firewall solution or a proxy server.
Ø Networks were made in accordance with the analysed strategies with different devices, topologies and protocol stack applicable.
Ø Server configuration will be chosen depending upon number of nodes and Network accesses, normally IBM servers (starts from basic configuration: IBM X servers).
For starting an IT corporate,
Ø The Hardware Components: Depending upon the business involved the number of computers and its configuration will be chosen.
(e.g.) if the corporate focuses on multimedia services, the configuration must be more supportive for VGA display cards (makes better video & audio effects) rather than processor configuration.
Ø The Software: As same of hardware components it depends upon the services, provided routine OS and packages (Ms-Office, Browser tools)
Ø Web / Software Development
H/w: Configuration with intel processor, with comparative RAM, HDD. Highly configured server (e.g. IBM BladeCenter HS21 XM).
S/w: Developing tools as JDK, VS2005, HTML tools, SQL server, Oracle
Ø Database Support (H/w and S/w will be same as development)
Ø Multimedia (H/w: basic configuration, supports VGA display cards as Intel, 810 GMCH, S/w: Adobe package, Animator tools, 3-D tools as maya).
Ø Matt Hayden, Teach your Networking (2nd edition), SAMS publishing, ISBN 0-672-32002-9
Ø Candace Leiden & Marshall Wilensky, TCP/IP for Dummies (5th edition), Wiley Publishing, ISBN 0-7645-1760-0
Ø D.Russel, The Principles of Computer Networking (1989), Cambridge University Press, ISBN 9780521339926