Any system that needs to handle vast amounts of data efficiently has to possess certain basic properties. These properties feed into defining the architectural principles, assumptions, capabilities, and patterns of the data-intensive system.
Let's discuss some of the core properties that we can find in a data-intensive system:
- Robust and fault-tolerant
Today's systems are distributed by nature. What that means is that different components of the system usually run on different machines and those machines could either be located together, or in an entirely different time zone.
One of the many, many reasons behind building such an architecture is so that each individual component can be executed on a much smaller piece of hardware, requiring far fewer resources to function. This being the era of commodity hardware, it means that these components usually run on commodity hardware. Given that commodity hardware is usually unreliable and can go down randomly...