RSA brings big data analytics to security threat management

Secure Analytics Unified Platform will let enterprises uncover threats much faster, company says

RSA has unveiled a new tool designed to let enterprises detect security threats more quickly than current technologies permit by combining big data management and analytics approaches with traditional network monitoring and threat detection.

The RSA Security Analytics Unified Platform is built around the company's existing NetWitness threat detection architecture. It lets companies capture and analyze massive amounts of structured and unstructured data to speed up threat detection.

The new platform is comprised of a data capture infrastructure and a separate security analytics warehouse. The capture component consists of network appliances for collecting, normalizing, and analyzing massive volumes of network packets and log data; an "analytics concentrator" that aggregates metadata from the appliances; and an analytics 'broker' that provides a single access point for running queries across multiple brokers.

The security analytics warehouse itself is Hadoop-based and allows companies to store and stage petabytes worth of structured and unstructured data. The warehouse supports long-term data archiving, forensics and analysis, says RSA, the security division of EMC.

Unlike traditional security incident and event management products that are log-centric in nature, RSA's Security Analytics platform supports full network packet capture and network session deconstruction as well, said Paul Stamp, director of product marketing at RSA.

It lets security administrators gather and look at petabytes worth data from multiple vantage points to uncover threats that are very hard to discover using existing security tools.

"This is a responsive technology. It's about new detection capabilities," Stamp said. It's about starting to narrow the gap between when an attack is detected and when the attack happened, he added.

RSA's technology is among an emerging class of security products attempting to use big data management and analytics approaches to address security problems. Others, including IBM, HP, Symantec and Trend Micro, are all already working on similar products.

Like big data analytics tools in other IT environments, the security products too are designed to let administrators run queries against extremely large and varied data sets to uncover threat patterns that would otherwise have remained hidden.

Many security analysts and practitioners say that the need for such tools is growing. They are convinced that unless there's a better way for security organizations to ingest and analyze the gigabytes and even terabytes worth of security log and network event data generated daily inside companies, there's no way to fight emerging threats.

Existing signature-based security tools fail to provide a complete picture of threats that may be lurking inside a network because they are fixed function and designed to look only at narrow set of parameters, said Jon Oltsik, an analyst with the Enterprise Security Group. "None of the tools can take in multiple data feeds and then give you the ability to query the data," to look for hidden threats, Oltsik said. Such a capability is crucial at a time when attacks are becoming increasingly sophisticated, targeted and hard to detect, he said.

Big data technologies such as Hadoop, MapReduce, Pig and Hive give companies the ability to dig in at a level they simply cannot achieve with traditional security tools, he said.

Scott Crawford, an analyst at Enterprise Management Associates, said the fact that a vast majority of companies these days do not even know when they are breached highlights the need for a more data-driven approach to security.

"We have been plagued by much blindness when it comes to threat awareness," Crawford said. "Most people are taking weeks if not more to discover a breach. We are not seeing what we have to see."

The biggest limitation with current security tools is that they depend on alerting rules and triggers that are based on what is already known, Crawford said. "You have to build rules predicated on what is known," about a threat to detect the threat.

Most current tools do not fare very well when it comes to dealing with unknown threats. They do not support the sort of querying that is possible with big data analytics technologies, he said.

Despite the benefits, there are some major caveats associated with big data analytics in the security realm. The biggest has to do with the lack of professionals familiar with Hadoop and related technologies. While integrated products such as the one from RSA this week mask a lot of the underlying complexity, they still require a certain degree of knowledge of big data analytics.

The companies that will benefit from such tools are most likely going to be very forward looking ones with the skills and the resources needed to pore through and analyze big data sets, Crawford said. "This will become pervasive over time. As these technologies become more widely adopted they will become more commonly integrated into security tools."

Jaikumar Vijayan covers data security and privacy issues, financial services security and e-voting for Computerworld. Follow Jaikumar on Twitter at @jaivijayan or subscribe to Jaikumar's RSS feed. His e-mail address is

See more by Jaikumar Vijayan on

Read more about big data in Computerworld's Big Data Topic Center.

Copyright © 2013 IDG Communications, Inc.

It’s time to break the ChatGPT habit
Shop Tech Products at Amazon