Network Visibility Architectures: One Size Does Not Fit All
The term "network visibility" has been in the networking space for ages, but has never been as important or relevant as it is today. New and exciting methods of automation -- whether virtualization, the cloud, IoT or even best practices like network segmentation -- tend to emphasize innovation over visibility. As such, networks develop blind spots that mask network problems and even faulty devices.
In this environment, understanding what visibility truly means is just as important as having it. Visibility is what allows IT to control and optimize the network, along with applications and IT services. Without it, organizational speed decreases, network problems take longer to resolve and security threats increase. Knowing how to measure what "good" looks like is critical in modern, complex networking environments. Determining the best strategy to accomplish this requires serious consideration.
Here are a few tips for making your decision.
End-to-end visibility architecture
An end-to-end visibility architecture lets you see across physical and virtual network elements, into cloud environments, and of course, into your network traffic. Achieving it requires a plan. Many organizations grow into their networks, adding on components, analytics, compliance and security, one appliance at a time. While it is possible to piecemeal visibility components along the way, the method can create its own blind spots while leading to unnecessary complexity and higher costs. Instead of waiting until a problem arises, it is important to annually reevaluate network visibility infrastructure and plan accordingly. Assess network segments to make sure they are being monitored -- look into virtual cloud monitoring for your public cloud resources and test your existing infrastructure with realistic high-volume traffic.
If done well, a strong visibility architecture will dramatically increase your visibility depth and breadth, whether it is physical, virtual, out-of-band or inline security. And depending on the solution, scaling doesn’t have to be a problem. VaaS can scale up or down, with cost based on consumption.
Setting up for immediate return
Visibility architectures typically yield immediate benefits that improve security, from reduced troubleshooting times to decreased network downtime. Realizing a return on investment (ROI) typically takes less than a year and can happen in as little as six months when visibility is applied effectively.
To start, there needs to be access to the proper data. This usually involves physical or virtual taps capable of being implemented at any switch point to access data from relevant segments of your network. When implemented correctly, it removes the bottleneck caused by limited access points (e.g., SPAN ports). And if you have planned well and overprovisioned your SPAN ports, it is recommended to intelligently aggregate them before getting them to your performance monitoring and compliance tools.
Next comes the filtering component to optimize the flow of relevant data to the right monitoring tools. Intelligent network packet brokers (NPBs) aggregate data, as well as filter, de-duplicate, time-stamp and even load balance the data sources to ensure monitoring tools are not overwhelmed. NPBs provide greater control while extending the life of existing network, application and security tools housed in the network -- especially for higher speed networks. When selecting a network packet broker, be sure it is the right size for your network.
Most NPBs have some level of context-aware data handling capabilities, though some perform at high speeds better than others. Having security intelligence awareness is very rare. By authenticating applications at high speeds, it is possible to send good traffic, like Netflix or Amazon Prime streams, for lower levels of security analysis. This is not possible if you cannot identify those applications at full network speed. Security intelligence can also identify what applications are running, the bandwidth used by each, the geolocation of application use, device types, routing of information and perform SSL decryption -- all within the visibility infrastructure layer.
Why it matters
Complex networks need good visibility to keep them safe. Besides the obvious, there are four typical and important use cases for a strong visibility architecture.
- Strengthening security defenses. If your network infrastructure is dropping packets at high network speeds, it is only a matter of time before an attacker will sneak in and exploit this weakness. Network and troubleshooting visibility needs to keep up with these high speeds and easily digestible insight into traffic sources and destinations.
- Network failure prevention. Prevention (not remediation) should be the goal on any network. The right visibility approach will have bypasses for inline devices, high availability failover paths that revert immediately, and can tell you which applications or network segments are underperforming.
- Faster time to repair. The two biggest fears of every IT manager is (1) being breached and (2) network downtime. Remediation of either of these is all about troubleshooting time, which is driven by your visibility layer. Having a properly sized, intelligent visibility layer can reduce mean time to repair by up to 80%.
- Test your network periodically. A good practice is periodic network testing, which requires capturing data for analysis. It makes isolating issues and resolving anomalies much easier and faster. To do this, you need to capture the data and have the ability to play it back within your network staging area. This kind of "traffic rewind" capability is simple to implement but can only be done with proper planning.
Ultimately, one-size does not fit all in terms of visibility architectures. You need one that is sized to your network’s specific configuration and needs. When done effectively, though, it can prevent and solve a lot of the daily problems faced by your network operations and security teams. Not all visibility architectures perform equally.
— Jeff Harris, Vice President, Solutions Marketing, Ixia
CALLING ALL TEST, ASSURANCE AND MONITORING COMPANIES: Make sure your company and services are listed free of charge at Testapedia, the comprehensive set of searchable databases covering the companies, products, industry organizations and people that are directly involved in defining and shaping the telecom test and measurement industry.