San Francisco-based Riverbed Technology is making Big Data moves this week. The company's latest innovation makes it possible for Riverbed Opnet AppInternals Xpert users to apply Big Data collection and analysis technologies to application performance management, also known as APM.
Riverbed's Big Data techniques give application support, developer, and QA teams insight into the end-user experience. The result, Riverbed said, is faster problem diagnosis, increased IT efficiency, and improved application service levels. Riverbed is using a combination of Transaction Trace Warehouse, a comprehensive record of all application transactions, and Big Data analytics to make it happen.
Alain Cohen, co-founder of Opnet Technologies and CTO at Riverbed Performance Management, said the full potential of APM lies in exploiting the information that sits in large volumes of detailed data about transaction performance and behavior.
"Our Big Data capabilities go far beyond traditional statistical performance summaries by enabling massive retention of transactional data and powerful unstructured search that gives IT the ability to quickly find and analyze the right set of transactions to definitively answer questions," Cohen said. "Equipped with these capabilities, IT organizations can rapidly detect performance problems and pinpoint their root causes before users or the business are impacted."
Overcoming Data Overload
Riverbed's solution aims at a challenge in IT departments: The scale , diversity, and granularity of data required to comprehensively manage application performance quickly overwhelms traditional performance management solutions. Redundant and virtualized infrastructure , load-balanced applications, and service-oriented architectures also contribute to the exploding volume of available performance data.
Given the volume and speed of available data, analytics are essential to identify potential threats and service disruptions. Historically, performance management solutions have dealt with this volume of data by sampling, averaging, or otherwise reducing the granularity of data they collect. Riverbed said these techniques leave customers with inadequate data and insight, losing the ability to track the complete application performance picture for each user and every application.
In contrast, Big Data collection and analysis techniques gather all data, however large and varied, into a single framework for analysis. Application support teams, development teams, and IT managers gain the ability to search, analyze, and respond to the specific challenges of specific users or individual transactions rather than just managing performance averages. This level of insight is often completely hidden by traditional application performance management tools. (continued...)