Get the Best Out of Your Network

Even with a well-planned and run network, there’s often the need to change configurations and traffic routing to keep the network running efficiently. Traffic-routing changes are costly and risky – especially when done manually – but are a necessity in most transmission networks.

Service providers wish for networks that are better at taking care of themselves; an example of this is given here in a plea for equipment-level features that would help in the efficient operation of WDM networks. Until that happens across all network technologies, operations and planning staff have to continue to take care of network efficiencies themselves. Until that happens, here’s a description of a handy feature of a good network-analysis system that can help you get the best out of your network and save you time and costs in dealing with routing inefficiencies:

Networks are mostly built as needs arise and by necessity grow one link at a time. As soon as a new node is commissioned, traffic gets put on it. Satisfying a new demand as quickly as possible is key for customer satisfaction and revenue collection.  Over time, more and more nodes are added, new links are added and existing links are upgraded. At some point, you install a new node that provides a direct path between nodes. This means that traffic between those nodes now has the option of taking a cheaper direct route, rather than being routed indirectly and consuming extra network resources. To make more efficient use of your network, you’d like to use the more direct path to liberate capacity for other traffic.

With the continual addition and deletion of services consuming paths in the network, you end up with fragmented use of the network facilities. For example, while half the channels in a TDM facility may be available, with the arbitrary addition and deletion of services over time, the available channels are unlikely to be contiguous.  Likewise for WDM facilities, wavelength usage gets fragmented across the available capacity. At some point, you get an order from a customer that requires a set of contiguous facilities and you don’t have them. The result is a provisioning delay for a high-value service which doesn’t do the customer or you any good. To make room for this new demand, you will need to undertake a set of manual service re-routings.  As mentioned above, these are costly and risky.  

What’s Wanted

This network defragmenter needs to be careful. The cost and risk of moving circuits needs to be balanced against the benefit of the defragmentation produced. You want to gain the biggest optimization/defragmentation benefit with the fewest routing changes.  Where routing changes are required, you want to know exactly where and how the changes need to be made.

A what-if analysis environment is required to show the cost savings using different metrics such as shortest distance, minimum hop, minimum cost against the possible groupings of circuits that can be moved and rerouted. Ideally, if this is run routinely and scheduled as normal preventative care in the network, the underlying time-critical problem of making space shouldn’t ever happen.

Russ Green is SVP, Product Management & Marketing, VPIsystems.  He has 15 years of enterprise software experience in large-scale, high availability systems, working with globally-distributed development groups and customers. Before joining VPIsystems, Russ was the vice president of Development for 724 Solutions, based in Switzerland.