NetBrain’s Network Data Model and the Foundation for Automation
IT’s Big Data When IT problems strike, the key to prompt resolution is hiding in the data – data produced when the fault occurs, historical data, and live data obtained…
August 30, 2018
We are familiar with manual network diagrams; these are the MS Visio maps we are accustomed to using when documenting or reviewing a network topology and configuration. You might also remember the road atlases of yesteryear – the thick spiral-bound books that were part of the family’s road trip kit. Today, when hitting the road, we no longer use atlases. We made the switch away from paper maps to online or on-board navigation systems over a decade ago! However, the MS Visio diagrams are still the de facto network map in many organizations. Somehow, when it comes to mapping our networks -the complex maze of virtual highways that deliver our applications and drive our businesses- we are still driving old school.
As Automation becomes more and more of a best-practice to manage the complex networks of today, we often hear the term Dynamic Map as the natural progression for manual diagrams. Like “Google maps” is to road atlases, network Dynamic Maps are the ideal visual vehicle for documenting and troubleshooting the modern network and moving away from the vintage static diagrams.
But, what are these Dynamic Maps? What exactly makes them “dynamic”?
Driven by Data
There is a monumental amount of data that feeds a Dynamic Map. In essence, it is a rendering of a mathematical model derived from live network data (device images, properties, config data, and more.). In the same way that Google, Bing or Yahoo gather data from many sources to update their interactive online maps (cars with cameras, user reviews, crowd-sourced traffic data, etc.), the data behind a network Dynamic Map is cleverly collected through several methods which extend beyond SNMP polling alone (e.g. telnet/SSH or REST API) for accuracy and completeness.
What makes a network map useful is the ability to provide data and context around a problem or task. A Dynamic Map enables the visualization of multiple layers of data. The user can zoom in and out of the map to reveal or hide the additional details, much like a car’s navigation system. In contrast to Visio diagrams, which can be overwhelmed by too much data, this feature of Dynamic Maps makes them less congested and easier to read.
The ability to toggle data views on or off the map gives the engineer an unlimited amount of data, in an easily digestible, measured way. In this manner, a Dynamic Map can provide insights into network performance, design, security and troubleshooting within the context of the map without information overload.
In an interactive online map, the user can view additional landmarks and points of interest by zooming in and out of the map. These points of interest will reveal additional details when clicked on.
In an Dynamic Map, the user can view layers of data by zooming in and out of the map. The user can drill into each device for additional details.
Third Party Data Integration
When a task requires data from multiple sources and tools, engineers often navigate through several windows and screens to gather it. In an environment like this, it’s easy to lose the context of the task in the process of juggling multiple troubleshooting tools, and it becomes difficult to correlate all of the data effectively.
A Dynamic Map can integrate seamlessly with existing workflows and tools, and serve as a shared console for a single, contextual view. If we borrow an analogy again from our car navigation example, this is not a different concept from the multiple data sources that an application like Waze might have. If you’ve ever used this popular collaborative navigation app, you realize that the richness of the application is due in part from the inputs gathered from users and third-party integrations.
In a Dynamic Map of the network, third-party integration helps engineers keep all the necessary data on the same map. This single, contextual view makes the correlation of information from network monitoring, trouble-ticketing, CMDB, event management systems much easier. API integration makes a Dynamic Map a truly “Single Pane of Glass.”
A single, contextual view makes the correlation of information from third party tools much easier. API integration makes a Dynamic Map a “Single Pane of Glass.
It Updates Automatically
It makes no sense to create the ideal map for a specific task only for it to be obsolete as soon as something in the network changes. To draw from our online map comparison, thanks to crowdsourcing and other live inputs, online maps are kept up-to-date, sometimes with the latest delay and road construction information!
So the beauty of the mathematical model behind Dynamic Maps comes from its interactive nature. As changes are carried out in the network, they are updated automatically on the map. This model allows the user to create maps from live network data, precisely at the moment they are needed, but most importantly, through recurring network benchmarks maps can be updated automatically every time they are opened.
Diagrams on Demand
Every IT workflow can be enhanced by the kind of visualization that a map brings to the equation. However, even the best-kept network diagram repositories seldom provide the one needed for visualizing a specific corner of the network.
For example, a slow application may traverse devices located in separate data centers. In this case, having multiple maps of the datacenters with irrelevant devices on it is not very helpful. Troubleshooting this issue requires a single map visualizing the entire application path, from source to destination, across data centers. In our road map example, our task is to drive home from work and avoid traffic. The map you need in this case is the best path from your current location to your home. Even better, this map can give us updates on traffic real time, on our route from work. Both of these examples require a straight-forward A to B map.
A network Dynamic Map offers the ability to generate a A/B path on demand, in seconds. In other words, the map adapts dynamically to the task at hand, helping to save the day.
Dynamic Network Overview
There are times when a site-wide or network-wide view is needed. Whether it be for a project-specific task or to gain a better understanding of the connections between sites, this global visibility of the network is valuable. A Dynamic Map can provide a single global view of the network and enable the network engineer to drill down, site-by-site.
Much like zooming into a Google Earth global map to understand borders, land features or points of interest, the user of a Dynamic Map can zoom into it to reveal details gathered over time. For instance, a complete review of the layer-3 topology and design data of a site is needed. From this global overview map, a Dynamic Map allows the engineer to zoom into the site, revealing the level of detail needed and the connections between other sites. The map should also be able to provide comparisons between the live network and prior benchmark data (much like how you can view a map over time in Google Earth.)
Understanding the advantages and benefits of Dynamic Mapping over static diagrams can help network engineers make an informed decision on selecting their next network automation solution.
From this global overview map, a Dynamic Map allows the user to zoom into the site, revealing additional levels of detail and the connections between other sites.
To evaluate whether you need a Dynamic Map, or an automated static diagram, identify if you need a better user interface for network automation or a better reference document of your network.
How long did your last outage or network degradation take to resolve? What did you spend the most time on?