Back at the dawn of time it was easy to spot a server. It was that big box in the data center into which everything else connected -, the central brain that controlled everything. You plugged in some “peripherals” for data storage (disks and tapes), and a few input and output devices, and you were in business. Human beings in the data center scheduled the jobs and deadlines were measured in hours (sometimes days!). In biological terms, we built a computer that worked like a mammal – like us, in fact.
What's the biggest difference between then and today? Reaction time!
Imagine a world where autonomous cars are controlled by a giant central brain (sounds like an old SciFi movie, right?). If a sensor in a car detects an obstacle, it would have to send that information back to the central brain, which would then calculate evasive action, and transmit the instructions back to the car …by which time the accident would already have happened!
Of course, autonomous car makers avoid this dilemma by allowing a small computer (server) in the car to make independent decisions without waiting for "Big Brother". By processing data locally, the car can react fact enough and (hopefully) avoid a crash.
But this principle doesn’t just apply to cars of the future. As businesses today engage in digital transformation they need to react ever faster to real-time data collected by a vast array of sensors and inputs – the so called “Internet of Things” - so much data, in fact, that we give many parts of our IT systems their own “brains” so they can react quickly without waiting for a central controller, or add value to the data locally (such as identifying a face in a video) so they communicate actionable information rather than just a stream of bits or pixels. These capabilities allow companies to scale economically, improve customer experience and develop new revenue streams that would otherwise be impossible.
In short, today’s infrastructure operates more like an octopus. “Why an octopus?”, you ask. Well, every octopus arm contains its own autonomous intelligence, smart enough to identify prey by touch and quickly grab it before the animal’s brain even has time to receive the nerve signal. This makes the octopus an efficient and formidable hunter.
At Hitachi Data Systems we use a similar approach when we build small independent servers into our solutions and appliances, where they often support specific functions, reacting faster and taking processing load off other parts of the system. These functions may include data ingestion, visualization of data in motion, speeding up distributed databases and many more.
Our small servers are based on industry standard technology, and are optimized for efficiency, density and low cost. Of course, most data centers still require a workhorse with high power and advanced capabilities (as I’ve blogged before), but the old adage “horses for courses” applies well here.
To learn more about the distributed servers that power many of Hitachi Data Systems solutions and appliances, check out this introductory video and then dive deeper with a 3D model which allows you to interact with the server and examine it from every angle – even look inside!
Think of these small servers as the intelligent arms of your data center octopus – ready to autonomously grab valuable data and process it before it can slip away!