Glassbeam IoT Analytics Cloud App Embraces Apache Spark

Mike Vizard
Slide Show

Five Ways Automation Speeds Up Big Data Deployments

There’s no doubt at this point that when it comes to the Internet of Things (IoT), analytics applications and Hadoop are joined at the hip. The challenge is processing all that data in a way that creates actual intelligence the organization can act on in real time.

To provide that capability, Glassbeam has announced that it has integrated its SCALAR cloud analytics service with Apache Spark, a real-time processing engine that plugs into Hadoop.

Glassbeam CEO Puneet Pandit says that batch processing of Big Data inherently limits the value that organizations can get out of that data. Pandit says organizations are investing in IoT to gain control over processes in real time. But Pandit contends that’s not going to happen unless the data being collected is processed and analyzed in real time.

With more organizations than ever relying on Hadoop to collect that data, Pandit says that Glassbeam is extending its reach to include integration with the Apache Spark engine that runs in-memory on top of Hadoop. The Glassbeam analytics application itself continues to run on top of an implementation of an Apache Cassandra NoSQL database, but collectively, Apache Spark and Cassandra enable the SCALAR service to process much larger IoT analytics applications in the cloud, says Pandit. The entire SCALAR service is then made accessible via RESTful application programming interfaces.

In general, Glassbeam is making a case for not only centralizing the collection of IoT data in the cloud, but also doing it in a way that enables analytics to be run in real time at scale. In fact, as IT organizations continue down the IoT path, many of them are about to discover that given the thirst for information in real time, the batch processing of analytics applications that were the hallmark of traditional data warehousing environments is simply not going to cut it anymore inside or out of the cloud.

Add Comment      Leave a comment on this blog post
Nov 24, 2014 12:06 PM Fiona McMeill Fiona McMeill  says:
Given the throughput of IoT data, it's great to see companies recognizing that the big data stores that such data generates need analysis based on where the data is. Another consideration is the analysis of the data before it is event stored, in the cloud or wherever else. Your readers may be interested to learn how such in-stream analytics can help dictate what is ultimately stored - in this section: How you can analyze streaming data here: Your comment that batch processing of Big Data inherently limits the value that organizations will get out of that data is absolutely true. And storage of such data is necessary for advanced analytical model development. And it's the execution of such advanced analysis in stream will help detect patterns that are worthy of storing in Hadoop. Reply

Post a comment





(Maximum characters: 1200). You have 1200 characters left.




Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.