← Back to use cases
Observability
Building a web application for network monitoring in heterogeneous environments
Who the client is
An internet service provider seeking to implement network monitoring in heterogeneous environments. It wanted its network operators to have a clear overview of network topology and services, detect anomalies and perform automated root cause analysis.
The challenge
The ISP's network is heterogeneous, and its operators needed one clear view of its topology and services. Anomalies had to surface as they happened, and the analysis of a failure had to point operators to its possible causes, so they could recover quickly.
What was at risk:
- Downtime from failures spotted late, and operators left to find the root cause of a problem by hand
What we did
- Provided a team of a network architect, network engineers, developers and UX designers, with expertise in networking, network monitoring and monitoring data visualization.
- Analyzed the requirements of the network environment to be monitored.
- Designed the application architecture.
- Built a component for real-time monitoring of changes in the BGP structure for the service layer, and a component that creates a network topology for LLDP and BGP.
- Used algorithmic analysis and different heuristics to show the potential root cause, one or many, of a network failure.
- Built the Proof of Concept of the application, with a UI that clearly visualizes network topology and anomalous components.
How it works
The application takes operators from what the network looks like to why something failed, in four parts.
- Topology A component creates the network topology from the LLDP and BGP protocols.
- Real-time monitoring A component monitors changes in the BGP structure for the service layer in real time, and an anomalous component is highlighted in the topology.
- Root cause analysis Algorithmic analysis and different heuristics process the collected data and show the potential root cause of a network failure, one or many, as a suggestion to network operators.
- Visualization The UI, designed to the newest usability standards, visualizes the network topology and its anomalous components.
Outcomes
Technical
Anomalies highlighted
The network is monitored in real time, and an anomalous component is highlighted in the network topology.
Business
Faster failure recovery
The application analyzes detected anomalies automatically and suggests the possible root cause to network operators.
Made possible
LLDP + BGP topology
Operators get a clear overview of network topology and services across a heterogeneous environment.
Tech & stack
Python
Rust
Ansible
- BGP
- LLDP
Anonymized example — fuller write-up available on request.
