Job descriptions that are generic

Job that demands skilled professionals with extensive experience in network administration. Candidates applying for the job must have hands-on experience on routers, switches and enterprise servers.

Day to day troubleshooting of common network problems, ability to handle help desk calls, interact with IT staff in remote locations and resolve WAN related problems.

Job descriptions that are specific about technology

Looking for a network engineer with experience in Cisco firewall products, Wan accelerators, F5 load balancing devices, Blue coat, MPLS etc. Must be at least a CCNP certified professional.

As you have read, there is no common job description to this profile. We recommend that you choose a specific area of networking such as security, switching or routing and increase your knowledge in that area. Also, consider taking up certifications in those areas and aim to become a specialist over a period of time.

A computer network may be any set of computers or devices connected to one another exchanging data and information. There are three main types of networks:

• A Local Area Network (LAN)- described as a small network constrained to a small geographic area.
• A Wide Area Network (WAN)- typically a larger network that covers a large geographic area.
• Wireless LAN & WAN- a more popular approach to computer networking that involves communication over small or large geographic areas using electromagnetic waves to communicate between nodes without the use of cables.

Networks can be viewed both physically and logically. A physical perspective involves location, the actual cables, and network elements such as routers, bridges, and application layer gateways to connect the physical media. A logical perspective involves subnets, or maps of the physical media. Many campuses of buildings may use a set of LAN cables in each building to create a common subnet, using virtual LAN (VLAN) technology.

There are also distinguishing characteristics of networks which may describe them as an intranet or extranet. An intranet is a community of interest under a private administration which requires special access by users, which are most commonly employees of an enterprise. Intranets do not necessarily need to be connected to the internet. An extranet is the extension of an intranet which allows secure communications to users outside of the intranet community.

A collision domain can be a single segment of Ethernet cable in shared-media Ethernet, or a single Ethernet hub in twisted-pair Ethernet, or even a whole network of hubs and repeaters.

If you are familiar with the layers of the OSI model you can easily determine what devices extend or break up collision domains.

In case you are not familiar with OSI model, here is what you need to know.

OSI Layer 1 Devices (hubs and repeaters) forward all data that is sent on the media and therefore extend collision domains.
OSI Layer 2 and Layer 3 Devices (switches and routers) break up collision domains.

More specifically it is an area of the computer network made up of all the computers and networking devices able to be reached by sending a frame to the data link layer broadcast address.

In more real scenario, after a packet is broadcasted, all the devices connected to the same network can receive the packet. For example, if 10 systems are connected to a hub and one of the system broadcasts a packet then all the systems connected to that hub will receive the broadcast packet.

Tip: Virtual LAN’s (also known as VLAN’s) can be used on switches to reduce the size of the broadcast domains and at the same time increase security. This is also known as breaking down of broadcast domains.

Unicast is a type of transmission in which information is sent from only one sender to only one receiver. In another words, Unicast transmission is between one-to-one nodes (involving two nodes only).

Examples of Unicast transmission are http, smtp, telnet, ssh, pop3 where the request for information is directed from one sender to only one receiver at the other end.

To understand this better, try to imagine a scenario where one computer’s request for the URL [http://www.google.com] is received by every other computer that’s connected to the Internet. This way Internet will be flooded with such requests and therefore there is a strong need for Unicast transmission on small and large networks. All Ethernet and IP networks support this type of transmission.

Unicast uses TCP for its communication and packet transmission.

For further info, refer to – IETF RFC

Broadcast

Broadcast is a type of transmission in which information is sent from just one computer but is received by all the computers connected to the network. This would mean that every time a computer or a node would transmit a packet of type ‘broadcast’, all the other computers will receive that information packet.

Examples of broadcast are many but one that would be best to understand here is a computer booting up and requesting for a IP address. In this case, the computer which is booting up and requesting for an IP address does not know which computer will be able to provide it with an IP address. It will then broadcast a request packet on the network which will be received by all the other computers but will be acknowledged by only the computer acting as a DHCP server. One other example would be that of the ARP (Address Resolution Protocol) which will broadcast the address resolution request to all other computers on the network.

Broadcast uses TCP for its communication and packet transmission.

For further info, refer to – IETF RFC

Multicast

Multicast is a very much different from Unicast and Multicast in definition and application as well. It is a type of transmission or communication in which there may be more than one senders and the information sent is meant for a set of receivers. Please note that sometimes information might not be directed towards any receiver at all.

Multicast operates on a different class of IP address (Class D IP Series). All the other computers electing to receive information in a multicasted network must be a part of Multicast IP based network. Also, unlike Unicast and Broadcast which uses TCP for its communication, Multicast uses UDP. This is only because TCP does not support Multicast mode of communication.

One good example of Multicast based network is video transmission network in which one computer needs to transmit video channel to a specific group of computers. This way when the other computers are also a part of this Multicast IP network, they will be able to receive same set of data at the same time. Multicast offers savings on bandwidth and is the preferred way of data communication when data is to transmitted to a set of computers.

For further info, refer to – IETF RFC

In data transmission, the data following the header are sometimes called the payload or body. It is vital that header composition follow a clear and unambiguous specification or format, to allow for parsing.

A packet consists of a header, which contains the information needed to get the packet from the source to the destination, and a data area, which contains the information of the user who caused the creation of the packet. A good analogy is to consider a packet to be like a letter; the header is like the envelope, and the data area is whatever the person puts inside the envelope.

The life of one connection will usually comprise a series of packets; in some network designs, they will not necessarily all be routed over the same path through the network.

Many networks do not provide protection against loss, duplication, or wrong delivery of packets. However, it is possible to layer a transport protocol on top of the packet service which can provide such protection; TCP is an example.

A datagram is a self-contained packet, one which contains enough information in the header to allow the network to forward it to the destination independently of previous or future datagrams. Thus, no setup is needed before a computer tries to send datagrams to a computer with which it has not previously communicated, unlike with virtual circuit protocols.

Some systems that use packets (e.g. ATM) require a connection setup stage before any packets can be sent, which is why this distinction is drawn between the two otherwise identical terms. Others, for instance the Internet Protocol, do not require prior setup, and there the two terms are used interchangeably.

The term frame is sometimes used to refer to a packet exactly as transmitted over the wire or radio.

1. Packet
2. Header
3. Unicast, Broadcast and Multicast
4. Broadcast Domain
5. Collision Domain