Beyond Analytics to Real-time Decisions
Companies and business leaders are increasingly realizing the advantages and opportunities provided by implementing real-time decision making and action on streams of data. Joe McKendrick, writing for the DBTA thought leadership series, sees the trend as well — and has the evidence to prove it. He notes that a Unisphere/Information Today, Inc. survey “found that … 57%” of data executives and professionals “are seeing a strong demand for real-time information”. The speed of real-time can enable innovative and competitive applications, such as real-time fraud detection, IoT, and even, “healthcare [where] the ability to monitor … data in real time can be a life-or-death matter.” Even for less dramatic uses, such as insight generation, real-time offers a lot. McKendrick observes that the survey “found that the longest running time for data queries tends to extend more than 5 minutes,” and some have to wait “more than 1 hour.” In the onset of the machine to machine (M2M) age, even 5 minutes is too long to wait for key insights.
However, speed is not everything. As McKendrick notes, “[m]oving information at real-time speeds … is just one part of the challenge”. It’s great to have data quickly, but it is essential to have the correct data. What most NoSQL solutions and others leave out is that immediate data consistency is just as important as speed: “[a]ny effort to achieve real-time data availability needs to go hand-in-hand with data quality efforts.” Eventual consistency just isn’t fast enough to keep up with M2M applications. Without immediate consistency, any real-time application will just do the wrong thing quickly.
For some applications, the number of devices, connections, and transactions is well known and manageable. However, for applications such as IoT, the number of devices and transactions will grow exponentially. As such, a traditional long-scale planning cycle will either leave you underprepared with thousands of angry customers, or over prepared with underutilized infrastructure. Scaling dynamically and easily with demands is key to real-time applications. For “in-memory users,” the dream of “integrat[ing] data from a growing number of sources” in real-time “is a reality.” By utilizing in-memory data solutions, businesses can develop real-time applications with elastic scalability.
With leading in-memory data technology, you can create industry leading applications for the real-time revolution. The best in-memory database will provide real-time speed, immediate consistency, and elastic scalability. With VoltDB, you can have all of these. Learn more about the power and capabilities of an in-memory database with our contribution to the DBTA thought leadership series, or download and try VoltDB for free.