By John Lytle
Requirements for data network resiliency have traditionally varied significantly by industry and even within individual enterprises. For example, a manufacturer could experience a 48-hour breakdown of a wide area network without significantly impacting production if the shop floor has output plans and schedules already in place. For currency traders in a bank, however, mere seconds of downtime could mean millions in lost revenue, so resiliency is obviously a top priority. At the same time, losing access to an ATM for an hour or two would not be a catastrophic event. In each of these environments, willingness to invest in network resiliency would vary accordingly, and would reflect the amount of revenue at risk should a network go down.
All of this is changing in today’s era of unified communications (UC). With data, voice and video traffic increasingly shared on a lone network, a single point of failure anywhere on the network can affect all other traffic on that network, regardless of whether it is low priority or critical. To use the aforementioned banking example, if a bank’s UC network fails and ATMs shut down, the currency traders are also likely to be affected, so not only is the data network lost, but the voice network is impacted as well. In other words, revenue-critical traffic on any part of a network makes the entire network revenue-critical. This emerging reality makes network resiliency an increasingly prominent priority for CIOs.
A Sea Change for Network Management Practices
Increased globalization and the growing imperative of inter-connectedness represent another sea change for network resiliency management practices. Consider a global enterprise with a major presence in North America and Europe and branch operations in Brazil and Nigeria. In this case, the remote operations not only expand the scope of the global network, but they present some complex choices. The technology, service provider and pricing options available in Brazil will likely be very different from what’s available in Nigeria, and both will surely be completely different from what’s available in New York or London. CIOs in such a situation face some difficult questions regarding cost and quality tradeoffs. For example, how important is it to the business for the networks in the remote operations to be as fast and reliable as those in major cities? How much additional cost is justified to optimize the speed and resiliency of the operations’ remote networks? How much latency is tolerable for the remote operations? Whatever the answers to these questions, insight into local carrier options, regulations and business practices is imperative.
CIOs managing global networks have long recognized that no single carrier is truly global; each carrier has different strengths and weaknesses related to specific regions or technologies. As a result, multi-sourced arrangements are a long-standing practice in the network space, as clients seek to leverage specific provider capabilities in particular regions.
One fundamental challenge of the multisourcing model has been to effectively manage disparate teams of providers with potentially conflicting agendas. This challenge has resulted in part from the manner in which contracts have traditionally been negotiated, whereby providers carefully define their specific areas of responsibility. As a result, when an incident occurs that falls outside that prescribed scope, providers have a financial incentive to argue that the incident is the responsibility of another provider—a dynamic that often leads to an unhealthy culture of finger-pointing and infighting.
To address this challenge, CIOs today are taking a more strategic approach to multisourcing governance in order to ensure collaboration between providers. A first step is to include collaborative language in the contract to define expectations for collaboration, problem solving and innovation. In addition, end-to-end service level agreements can eliminate the “It’s not my job” syndrome by establishing shared ownership and responsibility (and risk and reward) for incident resolution. Finally, mandated governance forums, whereby all providers regularly convene to discuss issues, can be an effective tool to build a culture of cooperation.
In addition, the emerging model of the “Request for Solution” (RFS) is often well-suited to address the sourcing challenges of today’s networks. As discussed in my previous CIO Insight article, under the RFS sourcing model, the customer broadly defines their operations and requirements and also solicits creative solutions from a pool of competing providers. To take the example of a global enterprise with branch operations in isolated geographies, the RFS can facilitate creative solutions on how to integrate disparate communications platforms with different technologies, standards and capabilities, and quite possibly lead to a single global integration partner who owns aligning all the other partners.
This more open-ended approach is encouraging the innovation needed in today’s global network environments.
About the Author
John Lytle is a director at the Information Services Group (ISG). To read his previous article for CIO Insight, “The Challenges of Managing Complex Networks,” click here.
