Part 2 will discuss Reliable & Manageable Scaling and Fault Isolation & Simplicity.

Reliable & Manageable Scaling

Scalability is a crucial characteristic that must be considered by network designers.
In addition, a network should also be reliable and manageable, which promotes stability.  A network may be capable of expanding to several thousand routers. However, if there is a failure of some sort that causes instability due to greater processing and CPU usage required to compensate, the design won’t be successful.
A scalable network that is a challenge to manage or troubleshoot is also unacceptable.
Network problems that arise would take an excessive amount of time to address and repair, negatively impacting an organization’s business goals. Thus, a scalable design must provide a manageable and reliable network.
Combining scalability, reliability, and manageability of a network is a challenge. As a network expands, it becomes more complex, making management, maintenance, and repair more difficult.

Fault Isolation & Simplicity

Providing fault isolation boundaries creates fail-stop behavior for the wanted fault model. Fail-stop prevents faulty behavior from propagating past a fault isolation boundary. Effective network design must prevent a failure in one domain from spreading to other domains within the network. This is essential for the avoidance of unneeded processing and delays throughout the whole network because of a device or link failure. In addition, this will result in a converging network that has more stability, reliability, and speed.
The above concept also promotes simplicity as a network architecture characteristic, which decreases complexity in operations. Typically, fault isolation stops the propagation of network fault information across several domains or network areas. This will lead to greater network stability and the minimization of recovery time after failures.
The negative consequences due to a link or network node failure may spread to all devices across a network, resulting in a large volume of unneeded processing. A properly designed network using the Fault Domains Principle with its logical separation will limit the negative impact to within one fault domain.
Alpha Engineering Business Support
 Choosing an expert computer technology-support provider is an important decision for your business. Alpha Engineering Associates has been a trusted partner of many companies in the Washington, Baltimore and Annapolis areas since 1990. For affordable client-focused network consulting services and solutions, please call Alpha Engineering Associates today at (410) 295-9500.