Enterprises looking to improve connectivity across the internet of things and other wide-area data ecosystems basically have two choices when it comes to fiber optics: purchasing leased line services from a telecommunications carrier or implementing its own network using dark fiber.
Both approaches have their advantages and disadvantages, but in terms of cost and flexibility many organizations are finding that utilizing dark fiber is an effective means of supporting critical and non-critical enterprise workloads.
According to Jerry Gilreath, senior director of IT engineering at Raging Wire Data Centers, the more the enterprise comes to rely on cloud and colocated data facilities, the more attractive dark fiber becomes. Not only does dark fiber provide high levels of scalability and flexibility, but it allows for greater network customization than most leased services and comes in at a lower cost as bandwidth needs grow. This could become critically important as the IoT gathers steam and the number of connected devices scales into the billions.
Still, he adds that it is best to do your homework when it comes to leveraging dark fiber. For one thing, a thorough mapping of your unique IT connectivity requirements is essential. Low-bandwidth, long-haul services are probably not cost-effective for dark fiber, considering the upfront hardware investment involved. As well, dark fiber is strictly point-to-point. If you require multiple end-points, it’s best to do a cost comparison between available lit services and multiple dark fiber paths.
The most common use case for dark fiber is high-capacity local or regional connectivity, says attorney C. Douglas Jarret of law firm Keller-Heckerman. Dedicated high-bandwidth (1 GbE or higher) on metro networks are fairly expensive on a cost/Mbps basis, even under special access service rates. And since fiber typically has a 20 to 30-year lifespan, derivable bandwidth will increase over time due to advancements in underlying transmission technologies.
This is primarily the reason why dark fiber is emerging as the go-to solution for cloud and data center interconnects (DCI), says Nokia’s Gary Holland. As the enterprise starts to extend and integrate its data environment across regional footprints, the need for reliable fiber infrastructure on a predictable cost basis grows. Dark fiber has become so prevalent in many locations that price competition is getting fierce, which brings down prices, and this also provides for a range of options when selecting routes and connectivity to main fiber trunk lines. Many providers now offer single- or dual-drop service to customer premises, with redundant pathways available to improve resiliency.
The enterprise should be aware that utilization of dark fiber increases the need for adequate link monitoring to ensure connectivity issues are identified and resolved before they impede performance. Key features to consider when managing dark fiber links include continuous in-service monitoring based on real-time data collection, plus passive demarcation to lower operating costs and enable monitoring under extreme environmental conditions. It also helps to employ a standard, open management interface to foster integration with operational support systems and geographical information systems.
As with any connectivity decision, application needs should be the driving force behind dark fiber deployments. Cost savings tend to be substantial over time when you light your own fiber, but this will be of little consequence if performance is inadequate for the desired business outcome.
But when applied correctly, dark fiber provides a readily available solution for a wide variety of enterprise needs, and will likely prove increasingly valuable as virtualized data ecosystems continue to spread across the globe.