Part 2: A Collision Course… SIEM and XDR
In my last blog, I talked about how extended detection and response (XDR) has the potential to fulfill an important promise—but in reality, confusion reigns. XDR as a category doesn’t work. At its core, it is an architecture, one that can revolutionize how we think about security operations. In this blog, we’ll compare SIEM and XDR and how they can work together.
The Plight of the SIEM
When we talk about security operations, we would be remiss not to talk about the SIEM.
The Security Information and Event Management (SIEM) platform has been around for decades. It has had its struggles in keeping up with its promises, the most prominent being the delivery of a “single pane of glass.” Despite falling short, it has become a staple of security operations centers (SOC) over time and, in fact, has been a force for good.
But the recent explosion of data driven by modern technologies and IT transformation are exposing its limitations.
The shortcomings of SIEMs within organizations don’t lie simply with the technology, as business variables play a vital role in the success of an organization’s security strategy.
Intending to drive higher value, SIEM vendors try to accommodate more data throughput, translating to higher license costs for customers. This strategy does not work in favor of security operations centers, which rely on visibility across the environment to detect and respond to threats. The more the telemetry, the better the visibility—but the higher the cost.
XDR: Lipstick on a …
Now enter XDR, the new kid on the block that promises to do all things the SIEM promised but fell short of delivering—extending detection and response and automation to drive efficiencies. But when you look behind the curtain, you start to see that these two trains are on a collision course.
The truth lies in the backend. Whether you are pursuing “Native” or “Open” XDR, you need a big data backend, you must be able to ingest raw logs as well as alerts from the technologies that make up the ecosystem at a massive scale. Recent acquisitions by traditional EDR players are a testament to this need. But simply attaching this backend does not deliver XDR capabilities and is tantamount to, as the saying goes, “putting lipstick on a pig.” This would mean the EDR product would ingest all the data that has been directed to the SIEM already. Wouldn’t this mean that the security operator now has duplicate data after paying for it twice? Once to the SIEM vendor and now to the EDR vendor? What value would the organization derive, other than increased complexity and cost, with such an arrangement?
The next piece of the puzzle is detection. Detection can manifest in three forms:
- Native detections that may occur in point technologies (think an AV solution detecting malware)
- A written rule to correlate data across raw logs
- Detections of detections (think of the grouping of similar alerts into an incident)
While certain products like an intrusion detection system (IDS), antivirus software, etc., can do their own (native) detections, they still need to be synthesized centrally to gain a singular picture. And an XDR platform must be able to accommodate all the data, at scale, across the ecosystem. But wait—if XDR is born out of the endpoint, why would it need to produce detections outside of it? If the endpoint protection that these vendors purvey was so great, why would you also need their EDR solutions? If endpoint solutions were the panacea for all things cyber, why are we continuing to invest in a defense-in-depth strategy that includes network, cloud, and email security, just to name a few?
To fulfill its promise, XDR must be more than “next-gen EDR.” It must be the aggregator of aggregators, the distiller of distillers, the view of all views. It can’t be beholden to the endpoint; it must be independent and treat every message, every alert with the right degree of confidence and aggregate those in a way that delivers meaningful value to the organization.
The Illusion of Choice
As an architectural construct, XDR has the potential to revolutionize the industry, but we must learn from the mistakes of our past and ensure that we are doing it right this time. It must be open and extensible across the technologies that help us secure our business, and it must have the ability to detect threats based on logs and other detections. Most importantly, it should consider the current reality that organizations will have data across and beyond their traditional environments.
The illusion is that we don’t have a choice. We need the promise of an XDR architecture to become a reality. We must be afforded the choice to do what is right for our business, the agility to grow with our business, and the capability to enable the growth of our business.
Learn more about how ReliaQuest combines XDR technology with services to make security possible for our clients. Our approach >