Improving Observability in Large Enterprise Networks with NetBox And SuzieQ

Modern networks are as complex as it gets in order to meet the needs of the market, which is placing ever greater demands on storage space, speed, flexibility and availability. However, they have never really been simple, as they are populated by very diverse devices that do not have a global view of the entire network. This design, together with the trend just mentioned, has meant that ever larger networks cannot be managed efficiently from a single privileged access point. We can list various "problems" in computer networks, such as: their complexity, opacity and unpredictability. Some concepts have emerged that aim to solve some of these problems: Observability and Automation. This thesis will focus on network observability, which is defined as the ability to understand the internal state of a complex system based on external outputs. When a system is observable, a user can identify the causes of a performance problem by looking at the data it produces, without additional testing and coding. In particular, we will analyze two tools that aim to make networks more observable: SuzieQ and NetBox. SuzieQ stands out for its ease of use, its performance and its ability to summarize heterogeneous information in a normalized "universal" format that is the same regardless of the input. NetBox strong suits are the completeness and granularity of the stored information and the great potential it offers in terms of integration with other tools. The graphical user interface is also worth mentioning: it is pleasant to use and complete, which is unusual in this field. The market is interested in integrating the two tools: they can complement each other, enrich the information provided to the user and mitigate some shortcomings. The aim of this work is to study their data models and then propose a solution to integrate the tools so that two main functions are offered: Synchronization and Validation. The main challenges lay in the architecture and code design, data normalization and defining the scope of the solution. I will outline all the phases of the study, the architecture of my solution and its implementation, the challenges I faced, how I overcame them and the reasons for my decisions. I will proceed step by step and, at the end, I will present the results from a qualitative – respected standards – and quantitative perspective – elapsed time measurements – as well as future work, as the core logic behind my solution will be part of a market product.