7 December 2015
One year on: Apache Spark Continues a Winning Battle Against Data Management Misery
In his first contribution to Compare the Cloud, Artyom Astafurov, Senior VP of IoT/M2M at DataArt, urges the readers to “take a deep breath and herald the arrival of Apache Spark – the missing link in data management set to make all our lives easier.” Following the development of structured data analytics tools to big data of today, comprised of sensor data from connected devices, social feeds and news streams, Astafurov points to the need of a streamlined approach to data management, suited to machine learning algorithms, streamlined analytic tools with faster performance and interactive approach to data exploration.
“In contrast to Hadoop’s implementation of MapReduce, Spark provides performance up to 100 times faster… Say goodbye to the days when we were forced to model data with one set of tools and then implement and run our models with another, and say hello to a more streamlined data management approach. Spark closes the gap between data discovery and running analytics in production, giving an all in one approach to looking at data by using state-of-the-art Functional Programming approach.”