Michael Bushong is the vice president of marketing at Plexxi.
Like everything from starting a business to building a house, true success doesn’t happen overnight. It can take years of planning to build a stable structure that will last more than just a few years. The same is true for your network.
Keeping up with the growing demands in today’s world of overloaded data centers requires tough conditioning so your network is in its best shape. Cisco’s 2013 Global Cloud Index report suggests that data center traffic will triple by 2017; 76 percent of that traffic is server to server traffic within the data center. With this in mind, many networks are already behind.
Revamping a data center network requires IT decision makers to step back and see the long-term potential by preparing for the growth and obstacles along the way. With one, three, and five year mile markers, consider this five year plan that every network team should apply to make sure their network can grow with demand in a linear fashion.
Year One: build a foundation for capacity
As companies build out their data centers, Year One is typically marked by an acceleration of capacity growth. The drivers of capacity growth have been well-documented: video and rich media driving higher volumes of traffic, mobility leading to an explosion in the number of devices and virtualization enabling higher compute utilization within the data center. Whatever the driver for a business, the result is an increasingly steep capacity growth ramp.
Given the predicted growth rate, the data center architect must add capacity across all resources. But more capacity requires more money, which can be limiting and is usually a derivative of the previous year’s budget. Companies that need to add capacity to meet this load demand must contend with the Year One problem year after year. The industry continues to expand and demands an efficient and capable data center. These contingencies rely on a solid, adaptable foundation.
Although this problem is recurring, having a foundation that can adjust to significant capacity changes will make it easier to address.
Year Three: asses and limit your operational expenses
After establishing a foundation that can adapt to increased capacity demand, the next step in Year Three is to determine how to handle looming operational expenses. The corporate gymnastics required to solve the recurring problem of capacity in Year One leaves IT leaders faced with an even more daunting challenge in Year Three. Even if IT vendors gave away equipment at no cost, the companies could not use it because of a lack of operational budget to manage the devices. For some companies, this issue in Year Three has already taken root. And while this situation may be uncommon, it speaks to the crippling effects of IT sprawl when operating expenses are left unmitigated.
Operational expense comes in various forms including environmental costs, management, and integration and orchestration. The fact that operating expense outpaces capital expense over the life of infrastructure has not been lost. Several approaches are designed to reduce operational cost:
- Fabric-based data center solutions
- Software-defined Networking (SDN)
- Development and Operations (DevOps)
The combination of options is unique to each network infrastructure. Don’t wait until the problem arises in Year Three to address it. Be proactive and solve this issue before it becomes one.
Year Five: scale your network
Now that you have a foundation to move past the recurring capacity problem in Year One and have enough leftover manpower and budget to address the longer-term operating costs from Year Three, we’ve reached Year Five. But imagine successfully navigating that gauntlet only to find out that it was all for naught. Further out on the horizon, beyond a point where most companies even consider real strategies or budgets, the Year Five problem looms. And that far out, Year Five has the biggest stakes.
Data center traffic is not just growing–it’s exploding. And architectural scaling goes well beyond merely providing capacity. If buyers aren’t certain that an architecture will ultimately prove scalable for large networks, they run the risk of outgrowing their chosen path. Such an outcome would be disastrous as future plans could require wholesale replacement of some or all existing gear.
The success in Year Five is contingent upon the building blocks in Year One and Year Three so be aware of this when making decisions with Year Five in mind. While some of the capabilities (or even platforms) might not be necessary during Year One and Year Three, any plan ought to consider long-term growth trajectories. Recognize that growth rates are only accelerating. The meaningful question for architects everywhere is how to transition from today’s reality to tomorrow’s promise.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.