The Journey to Next-Level Network Performance Management

The concept of observability or assessing the internal state of a system based on the data it produces dates back over half a century to the earliest days of computer science when algorithms based on control theory were first used to interpret system outputs. In the decades since, observability has become a key objective for SecOps teams, network engineers, and network architects seeking to understand and interpret the state of complex, multi-layered IT infrastructure.

With so much data traversing dynamic cloud and hybrid cloud environments, how much can you really know about the state of your network, and how fast can you react to problems? Too often, the answers to these important questions have been “not enough” and “too slow”. Fortunately, high-fidelity real-time data, vastly improved automation, and enhanced analysis capabilities are quickly putting next-level performance management and precision visibility within our reach.

Monitoring is Just the Beginning

Is improved visibility simply the result of more monitoring? While it may seem counter-intuitive, observability often results from distilling fewer (more essential) metrics to simplify data analysis while providing enhanced clarity and actionable insights.

• Monitoring is a passive surveillance and data collection process used to assess end-to-end metrics such as utilization and throughput, then compare the results to predefined KPI standards and alarm thresholds.
• Observability practices proactively delve into the behavior of specific devices, services or containers to characterize overall network health and the end-user experience.

Network monitoring tells you what, where and when anomalies like spotty connectivity or downtime are occurring, while next-level performance management practices decode interdependencies to tell you precisely why the end user is experiencing degraded service so that issues can be resolved quickly.

Bridging the Visibility Gaps

Unlike network monitoring, observability is a property of the system rather than a set of actions. This makes the journey to next-level performance an iterative process where the weaknesses of current methods are continually reviewed. New tools and features help to improve visibility and response times.

Deep Packet Inspection (DPI)

The limitations of conventional network monitoring tools lead too many applications and protocols traversing the network to be classified as simply HTTPS/HTTP, TCP or “other.” A deeper level of visibility differentiates applications like Outlook and SharePoint from low-priority streaming traffic, or worse, malicious applications that jeopardize network security.

Deep packet inspection (DPI) represents a quantum leap forward, with the information contained within packets providing an ultimate source of network truth. Advanced packet capture and analysis capabilities overcome monitoring limitations, allowing network engineers to improve decision-making and security.

Unified Communications Call Details

Unified Communications (UC) platforms must establish a reliable and consistent user experience across integrated voice, video and messaging channels. The increased reliance on UC means there is no time to waste when diagnosing UC call degradation issues. Traditionally, troubleshooting complex or intermittent issues required time-consuming packet-based analysis, perhaps even manually comparing multiple traces from multiple locations along a path. UC visibility has reached a new plateau through the seamless integration of packet and flow data. Actionable visualizations allow network and UC engineers to trace the path of a single call throughout the entire network infrastructure with visibility into the sources of degradation along that journey.

End User Experience (EUE) Scoring

Numeric end-user experience (EUE) scores distill the KPIs most essential to user satisfaction into a consistent, actionable format. This machine learning-driven scoring prowess allows network engineers and operations teams to quickly identify, prioritize and resolve performance problems by taking a deep dive into the behaviors of individual domains and unravelling in the behaviors of the client, network, server or application.

Certificate Analysis

Digital certificates and cipher suites facilitate secure connections and information exchanges between the client and server, but real-time status can be difficult to monitor and maintain. This lack of visibility can lead to compliance, security and customer support issues when system reliability is undermined by expired certificates or undesired cipher suites.

As part of a comprehensive approach to next-level performance management, certificate usage audits and expiration checks can now be streamlined and combined into a single interface to safeguard the network and brand reputation. Customizable dashboards and alarm settings provide the precision visibility needed to prevent the negative outcomes associated with non-compliant or expired certificates.

Microsoft Azure NSG Flow Logs

As applications migrated from the data center to the cloud, packets and VPC Flow Logs provided welcome visibility into AWS cloud environments.  As Microsoft does not natively support packet extraction, flow visibility is essential for troubleshooting issues within complex Azure environments. In some cases, these gaps can be filled effectively through the ingestion of network security group (NSG) flow logs. Harnessing this rich data source to characterize network traffic and behavior is an important driver of improved cloud and hybrid cloud visibility.

Fewer Metrics. More Clarity.

The next generation of network performance monitoring and diagnostics (NPMD) tools is removing obstacles that have lined the path to observability for decades, but there is still a long way to go. Just as the available tools and practices improve, multi-cloud environments, new cloud and edge-based technologies, and the ever-increasing volume and mix of traffic set the bar higher.

The future will see a continued shift from passive end-to-end monitoring towards proactive deep dives that will provide us with the what, where, how and why before we even ask the question. With a boost from AI, machine learning and automation technologies, peeling the next layer of the onion will be faster and easier.

VIAVI has leveraged decades of experience and collaboration to establish the industry’s most innovative and comprehensive NPMD solution. With the addition of enhanced DPI-based application identification, digital certificate analysis, interactive call visualizations and Azure NSG flow log analysis, the VIAVI Observer Platform brings the power of precision visibility and intelligence tog

ether to eliminate network blind spots. At the same time, enhanced EUE analysis and scoring allows virtually the entire IT team to become more efficient in problem identification, prioritization and remediation.

Level up your performance management today: www.viavisolutions.com/levelup