Network Configuration: a new source of observability

Have you ever considered Network Configuration Observability? For years, network observability has revolved around a familiar set of data sources: metrics, logs, events, traces, SNMP polling, streaming telemetry, and network traffic analysis. Each of these contributes a different perspective on the health and behavior of the infrastructure. 

But an interesting question is beginning to emerge:

How far can observability go?

Increasingly, organizations are asking not only how to collect more telemetry, but how to leverage every piece of operational information available across their infrastructure. The objective is no longer to monitor individual devices; it is to understand the complete state of the network from every possible angle. Each second the network is not performing properly represents a significant cost to the company

The trend is clear: modern observability platforms are expected to ingest and correlate virtually any log, metric, event, or operational dataset generated by any vendor across the network.

From Monitoring Platform to Unified Operations Platform

As Viewtinet has evolved over the years, many of our customers have gradually replaced traditional monitoring tools and standalone Network Management Systems (NMS) with a single unified observability platform. The goal has never been simply to consolidate dashboards. Instead, it has been to provide a single platform capable of collecting, analyzing, correlating, and understanding every relevant source of operational data.

This philosophy recently led us to introduce Network Configuration Manager, extending Viewtinet beyond traditional observability while maintaining the same vision: one platform for monitoring, management, analytics, and AI-driven operations.

Network Configuration Management Is Nothing New

The capabilities of a Network Configuration Manager are well established.

Network administrators rely on these platforms to:

• Automatically back up device configurations
• Restore previous configurations
• Deploy configuration changes
• Compare configuration versions
• Track configuration history
• Detect unauthorized modifications
• Support compliance and auditing processes

These capabilities remain essential for day-to-day network operations. However, the real opportunity lies elsewhere.

Configuration Data Is Also Observability Data

Once configuration management becomes part of a unified observability platform, network configurations stop being just administrative files. They become another operational data source.

Just as logs describe what happened, metrics describe performance, and traffic analysis explains network behavior, configurations describe how the infrastructure is intended to operate. This contextual information becomes incredibly valuable when analyzed alongside every other telemetry source. A configuration change is no longer an isolated administrative event. It becomes another observable event that can be correlated with alarms, performance degradation, application behavior, traffic anomalies, and infrastructure events.

AI Needs Context

One of the biggest challenges for AI-powered observability is not building better algorithms.

It is providing AI with sufficient context.

A machine learning model or a Large Language Model can only reason about the information available to it. The richer and more diverse the available datasets, the more accurate the conclusions become.

This is precisely where configuration data plays a critical role.

Within Viewtinet, AI-driven Root Cause Analysis continuously analyzes all available information sources—including metrics, logs, events, topology, network traffic, and now configuration history.

When a network issue appears shortly after a configuration modification, the platform can automatically correlate both events without requiring manual investigation.

Instead of an engineer spending hours asking questions such as:

• What changed?
• Who modified this router?
• Was a firewall policy updated?
• Did someone alter the routing configuration yesterday?

The platform can identify those relationships automatically as part of the RCA process. Configuration history becomes evidence. Not just documentation.

From Management Feature to Intelligence Source

Interestingly, Network Configuration Manager was initially introduced to address traditional NMS use cases. Indeed, backups, version control, compliance, and configuration deployment remain indispensable capabilities.

But once configuration data is integrated into a multidimensional observability platform, its value changes completely: it becomes another dimension of observability. Another source of evidence. Another dataset that AI can correlate with thousands of other signals to understand not only what happened, but why it happened.

As observability continues to evolve, the boundary between monitoring and management is becoming increasingly blurred. Every operational dataset has the potential to contribute to a better understanding of the infrastructure. Configuration files are no exception.  They are no longer just backups waiting to be restored. They have become one of the most valuable sources of contextual information for modern AI-driven observability.