Network Stability

Network Stability

The networks of today are much more stable than they were 10 years ago. As bandwidth capacity has continued to increase, so has the amount of memory and hardware capacity of the devices (routers and switches) that direct traffic on private and public networks. With this increase in network capacity and network element hardware, packet loss is far less noticeable in the enterprise network and very rare across the public Internet, primarily because the bottlenecks have shifted away from the network.

Network stability is still a key factor in today's networks as some networks transition from Fast Ethernet (100 Mbps) to Gigabit Ethernet (1000 Mbps), or even to 10-Gigabit Ethernet (10,000 Mbps). Traffic at these speeds will traverse a router or switch that needs to accommodate not just the traffic of one client to one server, but potentially from hundreds or thousands of nodes.

Several components impact overall stability of the network, including each of the network elements (routers, switches, firewalls), wireless access points, cabling, and more. For outsourced or managed WAN environments, network stability also includes a service provider plan and its network offerings that connect the data center and remote branch location, including any optional redundancy or service level agreements (SLA).

An unreliable network has an impact on more than just the data that traverses the network. Factors that cannot be as easily monitored or calculated, such as employee morale, can cost an enterprise in ways that cannot be forecasted. Employees who depend on an unreliable network may avoid certain applications, or look to alternative methods to get a job done if they do not feel that they can efficiently or, more importantly, reliably perform their required daily tasks. In some cases, the employees may just refuse to do their jobs, because they do not feel that they have adequate tools available to complete their assigned tasks. As mentioned in Chapter 1, sometimes the application is to blame for poor performance, and in some cases the network is to blame. In this case, the network is to blame if stability is impacted to the point that users cannot get to the tools they need to be productive and drive revenue.

As an example, consider how your mobile wireless provider's network impacts the usage patterns of your personal mobile phone. Your mobile phone is dependent on an accessible network to allow you to make or receive calls. In your daily travels, you will find over the course of time that certain areas have unreliable cellular coverage. During your travels through those areas, you will disconnect from the cellular network either intentionally prior to the service disruption or unintentionally when you unknowingly enter the poor service area. Data networks and applications share many similarities. If your daily responsibilities call for you to access several applications, one of which has extremely poor performance, you will approach the usage of that one application differently than you do the others. If that one application resides in a data center that is a long distance away over an oversubscribed and unreliable network, then both the application's and employee's performance will suffer.