Can DOD networks handle the stress of big data?
- By Sean Applegate
- Jan 29, 2015
Defense Department networks are feeling a little stressed out.
In the last few years they’ve been asked to move more information across longer distances to meet mission requirements. The underlying trends that are driving this transformation – big data and consolidation – are both extremely strategic and necessary, but that doesn’t mean networks aren’t looking for some relief.
DOD is collecting, analyzing, and sharing more data than at any time in history. Sensors and soldiers are gathering enormous volumes of real-time intelligence from across the globe, and warfighters are increasingly reliant on bandwidth-intensive applications to access mission-critical information. The challenge is that this data explosion is happening against the backdrop of data center and application consolidation. Centralizing IT resources can bring significant cost and efficiency gains, but it also means that applications are traveling longer distances across networks to reach the soldiers that rely upon them every day. Essential cloud-based applications, like Defense Information Systems Agency's Enterprise Email or the Army’s emerging HR system, must be available to personnel spread all across the globe, and achieving efficient and effective access involves a few simple steps that DOD organizations can implement to rise to the challenge.
Gain end-to-end situational awareness
Like any strategic initiative, you have to understand your environment before you can improve it. When it comes to networks and applications, this requires full visibility into what apps are on the network, who is using them, and how they are performing. DOD leaders can utilize “application aware” network performance management tools to gain this critical insight - mapping their environment to pinpoint problem areas and enabling traditionally disparate teams to collaborate through a single pane of glass.
During consolidation initiatives, this visibility is essential for ensuring continuity of operations as applications are moved to other data centers. These aren’t physical assets that one can just up and relocate without repercussions. It’s a complex spider web of interconnected resources, and the ability to monitor performance before, during, and after a migration ensures organizations follow a very important rule: When you move it, don’t break it.
The reality is that you can’t fix, or improve, what you don't understand, and clear visibility into your end-to-end performance is an essential foundation that should be in place before you attempt to optimize performance to better meet your mission objectives.
Optimize for mission velocity
Once you’ve achieved visibility into your infrastructure, and identified performance constraints, it’s time to start optimizing. Typically, the three most limiting IT challenges for global applications are bandwidth, latency and errors. In DOD, it is especially challenging to deliver consolidated applications to remote teams dispersed around the world, often connected via satellite or tactical communications. Solving tactical performance issues isn't as easy as just buying more bandwidth, because it’s either really expensive or not available at all! Often the best, and most cost-effective, option is optimizing the application and protocols, so they are smaller and faster. This approach mitigates the effects of congestion, latency and loss. Compression, caching, protocol accelerators, application accelerators and WAN optimizers are common mechanisms to achieve this objective.
Optimized infrastructure often moves data two to 20 times faster than non-optimized infrastructure. This ensures that soldiers are always able to access the applications and IT resources they need to meet mission requirements. Think of it this way: An optimized network versus a non-optimized network is like flying five hours from New York to Los Angeles, instead of driving your car for 40 hours. Like flying, once you get used to optimized performance you can’t imagine living without it.
Achieve control and predictability
After you’ve achieved visibility and optimization, the next step is control. Some of the world’s most sensitive information travels across DOD networks, and security is imperative, but so is predictability. Continuous improvement processes help DOD leaders eliminate network errors and ensure consistent end-user experiences. Soldiers need to be able to count on applications to deliver predicable outcomes when performing essential tasks to ensure the safety and effectiveness of their unit.
The other important “control” distinction is the ability to differentiate between mission-critical traffic and other types of traffic – say, for instance, a command and control (C2) system versus soldiers streaming the Super Bowl while stationed overseas. The ability to use Quality of Service to differentiate and prioritize bandwidth resources between applications is critical. This control enables DOD IT leaders to protect limited bandwidth to assure mission-centric operations, while still making sure everyone catches the big game!
Consolidation is already driving enormous efficiencies across DOD, and the big data explosion is providing our military with intelligence that has transformed our mission effectiveness. Now we have to make sure our end-to-end infrastructure continues to improve and evolve, so that or service men and women can achieve optimal mission velocity.
Sean Applegate is director of federal technology strategy and advanced solutions at Riverbed Technology.