With so many indicators (KPIs) and metrics dedicated to monitoring security and performance across networks, clouds, clients, and applications, it is clear that this complexity makes it difficult for SecOps and NetOps teams to manage information overload and extract meaningful insights from so much noise.
VIAVI Solutions takes a different approach to eliminate noise and simplify problem resolution for engineers. An example of this in the Observer solution is the End User Experience (EUE) Score, developed with machine learning, or as Gartner describes it: Artificial Intelligence for IT Operations (AIOps).
Operations teams need to quickly understand which users are negatively affected by degraded services and, more importantly, where the root cause of the problem resides. However, they do not need an excess of red indicators with intermittent false positives (also known as “noise”), nor continuous uncertainty about the absence of real service problems (false negatives). This is where Observer’s EUE Score helps
The EUE Score analyzes each network conversation in real time by leveraging sophisticated analysis, automatically learning the unique characteristics of your environment, and adjusting scores to reflect what the actual user is experiencing. Real problems are detected instantly and noise is silenced. To show how the EUE Score assessment improves when machine learning is applied in an application and IT environment, let’s look at an example.
The machine learning function was temporarily disabled in the following screenshot to illustrate the EUE Score without machine learning. In the second image, machine learning has been enabled to provide a more accurate view of user experience.
Figure 1 illustrates an encrypted conversation with an EUE Score of 3.0 (critical) and a server-linked domain problem indication. It is important to note that final EUE scoring algorithms do not require access to encrypted payload data. The associated ladder diagram on the right is a detailed visualization of the conversations. After performing a more detailed analysis of the specific conditions of this application environment, the performance issues did not affect user experience to the degree indicated by the score in Figure 1. While the score should have reflected degraded performance, it did not warrant a critical score (false positive). In Figure 2, machine learning has been enabled and accurately assesses the impact of performance degradation on user experience.
Figure 1 – EUE Score without machine learning
- Are there problems in the network, server, or application?
- Where is it located?
- What is the resolution?
Figure 2 – EUE Score using machine learning
Additionally, Observer’s APEX interface, shown in Figure 3, enables visualization of a score for physical locations with the help of a georeferenced map. It also shows the trend of services, applications, or network devices.
The Observer solution is not only useful for diagnosing installed networks but also for evaluating new deployments of services, applications, platforms, and any system that operates over the network.
Figure 3 – Observer Apex with trend indicators, maps, and EUE Scores
This is what Observer’s End User Experience Score offers: visibility of real problems and how to resolve them. Millions of network conversations can be automatically analyzed and scored, focusing operations teams ONLY on those transactions where there is a problem. Why be distracted by noise when you can stop it? Learn more about the EUE Score in this solution brief. Then contact TECNOUS, VIAVI Elite partner, for a demonstration of end user experience in your own network.
[1] Lerner, Andrew. (2017, August 9). AIOps Platforms [Gartner Blog Network]. Retrieved 2019, April 3, from https://blogs.gartner.com/andrew-lerner/2017/08/09/aiops-platforms/
https://www.viavisolutions.com/en-us/literature/end-user-experience-score-brochures-en.pdf