Design is a key component of a network.
Network design should be a team effort, says Naveen Patil, senior vice president of the Networks, Information, Technology and Entertainment Group (NITE).
“You want the best of both worlds,” he says.
The key is to ensure the network is resilient to disruption.
“You have to be able to adapt to change,” says Patil.
For example, if a disaster strikes, there is a good chance the network can withstand the disruptions and remain functional.
And if a power failure occurs, there will be no outage.
This is particularly important if a network is not capable of keeping up with demand and it is being used by large numbers of people.
“The question is how many of them?”
For instance, if there is only a single customer using the network, there could be as many as 50,000 people using the system.
The answer to this question depends on what you are designing and what you need to do to ensure it works.
“In terms of redundancy, we need to have a set of policies to manage these different types of systems,” says Michael J. Pender, senior associate professor of IT and network management at Northeastern University.
“We need to ensure that these systems are designed to handle the type of disruption that could occur.”
If the network has to be re-configured every day, it will not function, and you will not be able do it on a daily basis.
“So that’s where redundancy comes in,” says Pender.
“With that being said, you have to understand how that redundancy is set up, how that network will work under the worst scenario.”
This redundancy is critical.
“It’s the same as a house that’s not in good repair, it needs to be designed so that the fire brigade can move in and put out a fire, and the police can go in and get that person out,” says Gautam Goyal, principal technologist at Cloud Computing and Data Center Solutions.
“What you need is a network design that can withstand all these scenarios.”
And you also need to consider what happens when things go wrong.
“There is a point where you have a major failure, and if you are not in a position to respond effectively, it’s going to create a situation where it’s no longer an ideal situation to be in,” Patil explains.
The next step is to design and implement network architecture and protocols to allow the network to cope with the different scenarios.
For a network to be resilient, Patil recommends ensuring that: Network switches are designed with redundancy in mind.
There should be redundancy in the switches that allow for redundancy, he says, to ensure they can withstand a catastrophic failure.
Network switches should be designed with enough capacity to handle a variety of network configurations, from one-to-many connections to the single-to‑many connections.