IT monitoring vendor ExtraHop Networks, Inc. announced a new way to connect its platform with open source big data...
hubs to expand IT operations analytics and optimization.
ExtraHop's platform captures and analyzes wire data to measure application and network performance and behavior. The company recently unveiled its Open Data Stream initiative, a framework that sends data directly into two open source database platforms -- MongoDB and ElasticSearch -- that are frequently used in big data analytics.
"Open Data Stream allows us to stream every transaction, or selected transactions, or selected transactional elements into a platform for multi-dimensional analysis," said Erik Giesa, ExtraHop's senior vice president for marketing and business development.
Enterprise networking teams have started looking for ways to share their operational data with their companies' big data initiatives, said Jim Frey, vice president of research for Enterprise Management Associates.
"There is some rich data available from the network perspective that can be valuable in these big data analytics efforts," Frey said. "[ExtraHop users can] get beyond traditional IT use cases and more toward optimizing IT operations and recognizing the business activity that's going through the IT infrastructure."
Network managers' efforts at integrating their platforms into a larger IT operations analytics initiative are often stymied because monitoring vendors do not allow customers to share the data they collect with third-party platforms, Giesa said.
That roadblock is breaking down, however, as enterprises demand help with analyzing data collected across multiple monitoring platforms. Many IT operations vendors -- including NetScout Systems Inc., Riverbed Technology Inc. and Corvil -- have started exploring ways to make the data that their platforms generate available for broader analytical efforts in an enterprise, Frey said. ExtraHop has previously integrated with VMware Log Insight and Splunk.
"[IT monitoring vendors are] all looking at this immensely rich data set, which is application awareness from the network perspective, and saying we need to find a way to make it easy for folks to tap into this," Frey said. "I believe this is a natural evolution of the [application and network performance management] sector, to move out of simply being a rich and highly functional network troubleshooting tool."
ExtraHop frees its data for IT operations analytics
An ExtraHop appliance typically does not store the raw wire data it collects. Instead, it analyzes the wire data it captures and stores only the results of that analysis for real-time reports on application and network performance. By feeding that same wire data into a big data hub, ExtraHop is making it available for further interpretation and correlation with other data sets within an enterprise.
"Big data is about combining different data sets into a common, open source data store so you can derive insights from that," Giesa said. "You can prove expected relationships, but you can also uncover new relationships and insights. [ExtraHop's Open Data Stream] is the first step for leading IT operational analytics toward a big data-like architecture."
With ExtraHop's big data connection, IT organizations have a variety of use cases open to them, said Jonah Kowall, research vice president at Gartner Inc., beginning with IT operations analytics. For instance, an IT organization could combine log data from all its servers, switches, routers and applications with wire data from ExtraHop and assemble a new reporting and analytics platform.
"I can search across those data stores and do problem isolation and problem determination more effectively than having all these separate tools," Kowall said. "The other use case it opens up is capacity planning and trending. Let's say I want to build some specialized type of analysis based on a business condition -- or if we're doing a data center move -- and I need to understand long-term trending of specific parts of the environment, you could do that much more effectively [with ExtraHop and big data analytics]."
IT organizations could also use ExtraHop's wire data feed to analyze how infrastructure supports or correlates with business transactions and revenue generation, Kowall said. "That's the ideal future, but not the typical use case. Being able to understand business transactions is very important, but it's typically a little too advanced for most IT organizations."