Our entire life is dominated by relationships. In the early days it’s your parents and siblings. As you grow older it’s school friends, then when you start work you grow your network and older relationships die away. Sometimes a relationship breaks down and needs mending, it sometimes takes a long time to fix and you probably should have focussed on it way before the problem got out of hand. Your relationships directly affect how and when things happen. Each relationship needs managing or it withers. Getting a good start in life is highly beneficial and allows you to leverage services and knowledge through those good relationships that you might not have access to otherwise. It’s the same in IT, applications and infrastructure start out small, then grow and change their relationship structure. Things are added and others removed, usually added! Things break or take too much resource if not managed properly. When it comes down to it, relationships matter in IT, just as in life.
The relationships that exist between an application and the underlying infrastructure dictates how an application is going to perform. The relationships dictate how fragile or robust the infrastructure is. They define the latency the application has to deal with and dictate what kind of scale the app can handle. Before the advent of API driven infrastructure, it was almost impossible to accurately know or create a comprehensive graph of the infrastructure relationships that exist – particularly across domains or platforms. Even if you had a relationship graph documented, when something changed your document was instantly out of date. Companies tried to create Content Management Databases (CMDB) to help, but the data quickly became stale and unreliable. People couldn’t trust the base data so, either the analysis based on it was wrong, or people just didn’t use the CMDB data in the first place. The good news is, now the data is accessible directly from infrastructure platforms via their APIs. End devices and applications send their telemetry and relationship data directly to their respective management platforms. It’s data you can trust and there’s a constant stream of it.
As an example of the move towards utilizing relationship data, large enterprises are creating massive graph databases that have every relationship type in the entire organization. To give you an idea of scale, think of a large bank. Banks have anywhere from 10,000 to 30,000 applications and each one of those has many touch points. They have upwards of 40,000 networking devices, over 100,000 compute devices and multiples of that number of Virtual machines. Some have over 200,000 employees and all the end user devices that need to support them. The number of relationships reaches over 1 Billion, yes, 1 Billion with a capital ‘B’! These ‘Data Lakes’ offer the promise of a single point of truth for the entire enterprise. However the scope of what you can do with this data is as massive as the amount of data available.
To utilize all the available data you need tools that are built for large scale, relationship based data sets. You also need something that is API native and allows you and your customers, both internal and external, to interact with the data easily and intuitively.
Hyperglance has the relationship graph as its core architecture. Hyperglance capitalizes on both APIs and more historic methods to also tap into more traditional platforms such as networking to pull in data from multiple sources. Hyperglance then automatically aggregates and correlates the data to give you a completely integrated topology graph, or map, displayed within your favorite browser using HTML5 and WebGL. Hyperglance models all dependency data collected, including dependencies across different platforms, to show any knock on effects from issues. We’ve all seen the tangle of nastiness that comprises a large scale topology map. To prevent this, Hyperglance puts its graph into an automatically laid out 3D environment so the user can rotate around and look ‘into’ the topology, not just ‘at’ the topology. Many other Hyperglance mechanisms also help make sense of the data such as: custom icons, collapsible and expandable grouping, and filtered viewing capabilities to name just a few.
An integrated topology map is useful in itself. But, the power of having a topology map increases tremendously by being able to truly interact with it for deeper analysis. Hyperglance overlays a wealth of useful data to analyze on demand. Every node, endpoint, and link has key attributes such as name, type, size and status to name a few. Alarming data is clearly viewable on the topology map with details available at the click of a mouse. Overlaying alarming data is especially useful for operational prioritization and easier troubleshooting. Metrics charting data is also a click away from the same screen which is very helpful for diagnosis and trending. Not only is the Hyperglance topology map rich with visual and analytical data, it takes interactivity a step further by providing the opportunity to take context-aware actions on the spot. Now, from the same topology view screen, you can see exactly where problems are, understand what other resources might be affected, and immediately troubleshoot resources via dashboard menu driven actions or directly connecting via SSH, RDP and the like.
To take the pain out of data collection and aggregation, Hyperglance includes out of the box integrations with VMware vSphere, AWS, OpenStack, Docker Swarm, Kubernetes, Nagios, and SNMP. You can be up and running with a topology map in minutes. Hyperglance also includes an open RESTful API to incorporate any data you wish for a dynamic topology map that suits your needs best.
Hyperglance believes that relationships are at the core of operational success. We have built an interactive relationship visualization framework where the only thing limiting you is your imagination and the data you can get hold of. Better understanding your relationships and making sure they stay strong for optimum success is now easier than ever – at least for IT.