At the Strata + Hadoop World 2017 conference today, MapR Technologies announced a smaller instance of its platform for processing Big Data analytics applications that has been optimized for Internet of Things (IoT) environments.
Dale Kim, senior director of industry solutions for MapR, says instead of trying to push all IoT data back to a local data center or into a public cloud service, MapR Edge makes it possible to deploy an instance of a converged data platform based on a distribution of Hadoop at the edge of the IoT network. That approach will not only cut down the amount of network bandwidth required to support an IoT application, Kim says it cuts down on the amount of time needed to generate actionable insights.
Analytics applications at the edge of the network, said Kim, will be able to interact with endpoints considerably faster, while continuing to share results of the analytics generated with backend systems.
“It creates a closed loop for generating insights,” says Kim.https://o1.qnsr.com/log/p.gif?;n=203;c=204663295;s=11915;x=7936;f=201904081034270;u=j;z=TIMESTAMP;a=20410779;e=i
MapR Edge is designed to be deployed in a three to five node cluster capable of storing up to 50TB of data using industry standard X86 processors. Those nodes in turn will most likely sit between IoT gateways at the edge of the network and backend systems residing in a data center or public cloud.
Whatever the approach, there’s clearly a rush on to push analytics as close as possible to the IoT edge. While there are likely to be multiple ways of accomplishing that goal, the one thing that is for certain is that the amount of analytics code that will soon be running at the edge of IoT networks will be considerable.