In August and September we introduced the concept of structured cabling, its importance and the types of cables that are used. The fastest, of course, are fiber optic cables that can transport more data than we can ever use.
In this concluding part, we will look at networking, standards, and some security concepts.
A Bit of History on Networks
Today, interoperability, and connection between computers are taken for granted. But it was not always so. Some 40 years ago, the concept of physically connecting computers to each other started in earnest in the labs of universities such as CalTech and large companies such as IBM, DEC, etc. The major breakthrough came from the Xerox labs at Palo Alto Research Center (PARC). Robert Metcalfe was asked to build a networking system to make it easy for everyone to use a printer. The issue was that PARC had hundreds of computers when the general norm was just one or two.
In 1976, Metcalfe and his assistant David Boggs published a paper titled, ‘Ethernet: Distributed Packet-Switching For Local Computer Networks’. In 1979, Metcalfe left Xerox to successfully convince IBM, Digital Equipment, Intel, and Xerox to work together to promote Ethernet as a standard. Today his efforts have made Ethernet an industry standard and the most widely installed network in the world.
Once of the first companies to successfully sell networking as a product cum service was Novell. In 1983, they introduced NetWare as a network operating system that had a star topology with a file server acting as a router for all data transfer. This supported the then popular operating systems – CP/M, as well as the new MS-DOS from Microsoft. In 1993, Novell kind of killed itself by introducing Network Directory Services that did away with the need for a file server. Microsoft introduced Active Directory, and this completely killed the need for Novell’s products or services.
Because NetWare used the concept of file sharing instead of disk sharing, it had 90% market share between 1980 and 1990s. It evolved over a period of time to running databases, email, web and other services.
What are Protocols?
Novell started the run of NetWare by creating it’s own protocol called IPX. Over a period of time, NetWare started supporting TCP/IP and Appletalk. Today TCP/IP is the global standard for networking and is used by the Internet also.
TCP/IP was born a little later after Metcalf published his paper on the Ethernet. In 1983, the Department of Defence wanted to connect some 400 odd computers across companies, government agencies, research bodies and other institutions. The network was called ‘ARPANet. Since the 400 odd computers were from different companies, the network had to reckon with multiple operating systems, and different file types. The easiest way was to create a transport mechanism that could just move data, and leave the file and data conversion to other software. Thus was born TCP/IP.
TCP/IP stands for Transport Control Protocol / Internet Protocol. It is a mechanism, using which, data is moved from one computer to another. TCP/IP has two major parts – TCP and IP.
The transport control protocol decides how data is moved from one place to another. It first accepts data of any size and type. It then breaks the data down to smaller chunks. These chunks are called segments, and the process is called segmentation. These segments are wrapped in a standard data structure called packets. The packet, in addition to the original data, also contains some information about the sender, the receiver, and something called checksum that is used by the receiver to validate the data received. If the validation fails, the receiver will instruct the sender to resend the data. If it succeeds, the receiver will ask for the next data packet. Once all the packets have arrived, it will be assembled into a duplicate of the original data that had left from the sender’s computer.
The Internet Protocol is like a traffic controller. Each packet that is sent has two host addresses – one of the sender and the other of the receiver. This is used by the gateway to make sure the packet reaches the right computer.
TCP/IP works on the basis of a client/server model. Every computer connected to the network is a host. A centralised machine that routes the data packet acts as a server. To eusure universal usage, once the data packet is in the network, it becomes stateless. The server identifies the packet, reads the address and, sends it to the actual receiver.
There are variations of TCP/IP that are used for specific purposes. All these use TCP/IP as a basic transport mechanism, but may do a very specific job. Some examples are HTTP that is used for browsing, FTP that is used for file transfer, Telnet that is used for remote connectivity, and SMTP that delivers your email.
OSI and Network Layers
All networking including Ethernet and TCP/IP have been developed on an OSI model that refers to Open Systems Interconnection. Essentially, OSI has seven layers – Physical, Data Link, Network, Transport, Session, Presentation, and Application. Each of these layers speaks only to a layer above or below it. Each of these layers executes a specific function. TCP/IP, Ethernet and other protocols work at various levels of OSI layers depending upon the need.
The Physical Layer is at the lowest level or first layer. This has all the physical aspects of networking such as cabling, connectors, modems, repeaters, routers, etc.
The Data Link Layer identifies computer on the network, organizes the physical layers into logical groups, detects and corrects errors, and controls the data flow.
The Network Layer’s main objective is to move data. In converts logical address to physical address, and takes the shortest path to the receiver.
The Transport Layer ensures the data reaches without errors. It works on the error correction algorithm built by the Data Link Layer. It also creates the data packets we spoke about above.
The Session Layer establishes and maintains connection between two nodes or hosts on a network. It re-establishes connection if gets broken for some reason. The Session Layer, because it determines who can talk to whom, is also called a ‘traffic cop’ of Networks.
The Presentation Layer acts as a translator between the application and the network. It converts data into a format that the network can understand, looks after data encryption, and manages passwords.
The final layer is the Application Layer. This does NOT refer to applications such as Word of Excel. Rather it refers to applications that provide services such as file transfer, file management, and email management.
As an organization or as a user, you don’t want others to have access to your emails or your data. What you have to protect are usability, reliability, integrity, and safety of you network.
At first, let us understand the threats and then see what needs to be done for various environments of usage.
There are a multitude of threats to any network. The most common are:
• Viruses, worms, malware
• Spyware and adware
• Zero-hour attacks
• Hacker attacks
• DOS or denial of service
• Data interception and data theft
• Identity theft
I have heard of these. Why would I have to be worried about these and how do I handle them?
For an organization, data is critical. You not only have your own data, you have data of your clients, vendors, employees as well as others. Your business partners expect you to safeguard what they gave you. You will get the trust of your business partners; will have better employee morale, and avoid costs that may be incurred if your business is disrupted.
At an individual level, all of us keep a lot of our personal stuff on the Internet, use the Net for financial and business transactions. If some steals your id, he could potentially siphon off your hard earned assets.
Security needs and ways to handle them are different depending upon who you are what your usage of the network is.
At a personal level, an anti-virus software, a firewall, and safe keeping of passwords are the basic security requirements. If you are using a wireless LAN, do not use the id that came with the router. Create your own. Have a robust password for your router. Do not broadcast the SSID of your router. Enable MAC address filtering so you know what computers are connected anytime. Use static IP address for your network devices.
For a small or medium business, first install a strong firewall and have a powerful anti-virus on all your computers. For logging-in, assign strong passwords that are changed regularly. Implement clear employee guidelines about usage of computers and network access. All employee log-ins and Net access must be monitored regularly and analysed on a frequent basis. Have a network analyser that warns your network administrator whenever a suspicious activity happens. If you have multiple offices, set up a VPN, or a Virtual Private Network. Have a strong backup policy for recovery of all data.
Larger companies would need all we described before, a very strong policy with punishment on the misuse of the network, and constant monitoring of the network. Multiple layers of physical security may also be implemented making it difficult for unauthorised people to go near your computers. Authentication can be dynamic with the system generating a new password for every log-in with the password being deactivated the minute you log-out, or even if you do not use the system for a certain time.
We do hope we have been able to enlighten you on networks, how they work, and how we have to be safe. Today a business cannot be run without computers or the Net. A few precautions will enable you to enjoy the experience and focus on your business.
For over 28 years now, Tri-Tel has been delivering advanced telecomminucation products including IP equipment and advanced cabling, paging, conferencing solutions as well as local and long distance Internet access. Tri-Tel services cover Chicago Metro, Northwest Indiana, Rockford, Elgin, and Schaumburg.