Cisco Blogs
Share

Enhancing Hyper-Distributed Analytics

- November 5, 2015 - 0 Comments

Today Cisco completed its acquisition of ParStream, a company that has created a specialized database built for the Internet of Things (IoT).

In hyper distributed data environments, massive amounts of data are being created and in a very distributed way. Many examples of this can be found in the IoT, from terabytes of data created by sensors places in offshore oil wells to extremely time sensitive data created by robots in manufacturing facilities. In these hyper distributed data environments, challenges arise with collecting, storing and analyzing data that can’t be solved with traditional solutions that rely on data to be in a central location before it can be used to derive meaningful insight. In some cases, there is simply too much data to move across the network, while in other cases the problem is that the data is extremely time sensitive and moving it to a data center or the cloud for action is just too slow. These cases require the capability to store and analyze data very close to where the data is created…near the edge of the network.

Slide1

This is exactly what Cisco and ParStream will do together.

Last year, Cisco introduced Connected Streaming Analytics to analyze data “in motion,” data moving across the network, and hadn’t yet been stored in a database. The ability to analyze streaming data means near real-time insight on data that has just been created. What Connected Streaming Analytics doesn’t do is store massive volumes of historical data, which is where ParStream becomes a perfect complement. ParStream can ingest and store massive volumes of high velocity of data, and allows terabytes of new and historical data in thousands of rows to be queried with sub-second response time, enabling better and faster analytics at the edge.

Together, ParStream and Cisco will help customers manage and gain insight from hyper distributed data at rest and streaming data in real-time. The footprint of ParStream’s database is extremely small – 20 times smaller than competitive solutions – in fact. That is significant, because it means that ParStream can be deployed on a much smaller hardware footprint, significantly lowering the expense of the solution. ParStream is also designed to manage centralized querying of hyper distributed instances of the database, making it easy for customers to query data in even the most massively distributed environments.

The technical specifications of ParStream’s database are truly impressive, and I can’t wait to start introducing Cisco’s customers to this incredible technology. Even more impressive than the technology, however, is the very talented ParStream team. I’m excited to welcome them to our Cisco Data & Analytics team, and know that the initial work we have done together is only the beginning of great things to come!

 

Join the Conversation

Follow @MikeFlannagan and @CiscoAnalytics on Twitter.

Learn More from My Colleagues

Check out the blogs of Mala Anand, Kevin Ott, and Bob Eve to learn more.

Tags:

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.

Share