Sergey Nivens - Fotolia
The internet of things -- or a growing network of sensors and devices that can transmit data without human interaction -- is coming at full force. In fact, 20.8 billion connected things will be in use globally by 2020, according to Gartner predictions. The internet of things produces a massive influx of data, which directly impacts the data center in a variety of ways.
Use the five tips below to understand how your data center strategy should adjust to the IoT challenge.
Plan carefully for an IoT project
While enterprises often begin business intelligence (BI) projects with a set goal, internet of things (IoT) projects are less straightforward.
IoT initiatives spur questions such as "which problems should my business address to better serve customers?" and require a different set of tools and experiences than more traditional BI projects. When it comes to IoT, start slow; don't rush to adopt new technologies. Before you integrate IoT, collect and analyze operational, machine and other types of data to assess the business' needs. In addition, remember that IoT projects typically demand a NoSQL databases, and require a new approach to data center storage and networks.
Since IoT projects are customer-facing, use lightweight and responsive monitoring and management tools.
Expect storage to be a big IoT challenge
As the popularity of big data and IoT projects increases, so does the amount of data that organizations gather. Since IoT data is different from traditional data, storage can be a challenge.
Depending on use, IoT projects require different types of storage systems. For example, they may need to ingest trillions of small files coming from sensor data. Alternatively, they might require high-bandwidth streaming of fewer but larger files from a continuous stream of images, such as surveillance footage. An object storage system is a viable option for IoT environments and, in many cases, will be compatible with Hadoop.
Secure your IoT deployment
Security is another common IoT challenge. As many devices and a larger volume of data enters the data center, new security threats emerge. End-to-end encryption can help to minimize most IoT security threats. Integrate encryption and key management functions with the IoT applications an enterprise uses.
IoT often includes the collection of private consumer data, so corporations should use authorization with IoT systems to ensure only approved personnel can access the information. Organizations will also need to ensure that data center admins can manage IoT data at a high level without the option to inspect individual files.
Evolve memory for IoT
Simply throwing more memory or processing power at an IoT deployment doesn't result in better performance; different IoT tools require different approaches to memory, and each approach has its own benefits and limitations. Before implementing the IoT platform, research each tool by finding out the resources it will require. Installing more memory than your tools can use, for example, will only waste power and heating resources without improving overall performance.
Explore edge computing for IoT
IoT necessitates a more efficient way to store and process the massive amount of data it accumulates. Edge computing is an option for that IoT challenge; it processes data closer to where it is generated, which produces faster results, requires less movement of data, reduces cost and increases security. Some IoT functions require data be processed and analyzed locally for viability, furthering the need for edge computing.
While there are many benefits to using edge computing in an IoT environment, some experts say it is unclear whether IoT gateways will be more expensive than a centralized cloud, especially for applications that don't need low latency and high security.
Learn how some companies use edge computing for IoT
How to implement an IoT strategy
Where is IoT data stored?