Detection, Prevention & the Single-Vendor SyndromeWhy security teams need to integrate 'Defense in Depth' principles into traditional solutions designed with integration and continuity in mind.
It’s a controversial statement, particularly when it comes from a security vendor, but it must be said: No single vendor can adequately protect your enterprise from all of today’s threats, let alone what might be on the horizon.
There is a misconception today that "complete prevention" is a realistic goal for an enterprise security program. As an adversary’s level of sophistication increases, the ability and efficacy of a single product at preventing arbitrary intrusions begins to decrease dramatically. As a result, security teams need to adopt a new mantra: Given sufficient time, motivation, and funding, a sufficiently capable adversary will find their way into your organization, regardless of the tools that you have deployed.
One can simply recall the tale of djbdns and how its touted impenetrable "secure code" written by cryptographer Daniel J. Bernstein, failed to stand up to focused scrutiny. The story, as reported by The Register in 2009, is a cautionary tale about assumptions that organizations continue to make about the vulnerabilities of many devices and applications running on the typical enterprise network.
Knowing that the development of a perfectly secure application, system, or device is effectively impossible, we must shift our focus from "prevention” towards "rapid detection, coordination, and response."
Regardless of market buzzwords, all vendors try to accomplish the same end goals by leveraging similar techniques and technologies. As discussed at this year’s BlackHat by Lidia Giuliano and Mike Spaulding (Lies, and Damn Lies), unsurprisingly, each vendor, based on their unique implementations, have certain strengths and weaknesses, depending on the type of attack or technique being tested.
Consider the following scenario:
- Product “A” may be able to detect attack “x”
- Product “B” may be able to detect behavior "y"
- Product "C" is unable to detect either "x" or "y," but is best positioned within the network to take the most effective action against the attacker
- As is typical in the modern enterprise, Product A, B, and C are managed by separate teams and do not share the same reporting or communication structure.
This is a scenario to which many enterprises are now being exposed to, generally for the first time during post-breach incident response and forensics. Tools that they may have had in place detected portions of the attacker’s activity, but none were able to combine their results together to take effective action against the intruder in a timely fashion.
As a result, security teams are beginning to realize that due to the overwhelming volume and increasing sophistication of the modern threat landscape, they must now combine the proverbial "Defense in Depth" principle with solutions that are designed with integration and continuity in mind. This means they need to prioritize partnering with companies whose cybersecurity offerings are open in nature and seamlessly augment security operations with automated detection, enforcement and remediation. Only in doing so can they maximize their chances of success against a motivated attacker.
Craig Dods is the Chief Architect for Security within Juniper Networks' Strategic Verticals. He currently maintains top-level industry certifications, holds multiple networking and security-related patents, and has disclosed multiple critical-level CVE's in a responsible ... View Full Bio