T1 Network

A T1 network is best described as a private network with T1 connections between locations. Each T1 network connection will likely terminate into a router at the customer end, giving the customer the ability to control and shape traffic as it enters or leaves the private network. Customers with more than two locations will need to make a decision on whether to design their network in either in a mesh configuration (see diagram) or in a home-run configuration. A mesh T1 network system will consist of multiple locations, all interconnected to each other.

In this configuration, a single link failure will not adversely affect the network, since location A could communicate through location B to reach a location C (where the link between location A and C failed). This can be accomplished through simple routing at each location, so that if packets were destined to location C from location A, location A's router would have route choices that may include A-C, A-B (and B has B-C as a choice), A-D (and D has D-C as a choice) and so on.

A fully meshed network for the T1 indicates that all locations have direct connections to every other location. Though a fully meshed network is the most expensive option from a network perspective, applications that require absolute "uptime" may require a fully meshed network for T1. If non-critical locations are a part of the network, a partially meshed network may suffice. In this example (right) the red link indicates a failure between location A and C. A could communicate to C via routing through B, D, or Main to restore communications. Simple

The Home-Run configuration for a T1's network best suits a

network with one main location and several "branch office" locations. In this case, the most critical site is the main locations, where all critical database and computing applications are located. Losing a branch office in this network only affects the branch office, not any other location. In this case, due to the reduced number of T1 connections, significant savings can be achieved in both network savings, as well as the required termination equipment. Network routing is also simplified in this configuration. The drawbacks of this configuration are two-fold: 1) A single link failure can isolate a branch office completely. 2) A main office outage will affect the entire network. This configuration has no built-in diversity. Each situation should be judged in terms of how critical a single application or location can be vs. the cost associated with moving more towards a partial or full mesh network.

Restoration is Important

As a result of a combination of leasing and building, carriers can offer diversity within their network. Before the days of SONET (synchronous optical network) technology, carriers struggled to recover from cable cuts, fondly referred to as "backhoe fade." Before 1998, when a cable was cut, two activities commenced at the same time, locating and splicing the cable, and manually rerouting backbone network traffic around the break area. Before more automated methods of re-routing traffic with batch commands for DS3-level Digital Cross Connect machines, the vast majority of traffic was never restored until the cable was spliced, hours later. For private network users with point-to-point circuits, this could be a severe impact upon business operations. SONET fiber speeds and "ring topology" offered the ability for networks to be built in a series of bi-directional rings, with each location having two physical route choices on a ring. If a cable was cut, traffic was simply re-routed in the opposite direction on the ring, typically within 30 milliseconds. At that time, no users were affected, and the fiber splicing crews could work with much less pressure.

Back to Top