Visualisation Method for Overlaying Sequence Diagrams on a Directed Graph
Publication Date: 2017-Jan-19
The IP.com Prior Art Database
This disclosure details a method to visualize large and complex data set in the Application Performance Management (APM) domain. APM data sets often have complex relationships that are hard to comprehend and this disclosure makes use of animation, 3D layout and various other associated techniques to highlight complex relationships in the data to the user. Relationships highlighted includes transient (temporal based) relations across a large path (ie indirectly related components).
Visualisation Method for Overlaying Sequence Diagrams on a Directed GraphIn Application Performance Management (APM), various middleware components and their constituent parts are often visualised as a Directed Graphs (DG). DG visualisation consists of nodes, represented by an enclosed shape such as a circle, and edges, represented by arrows, which link the nodes together. Nodes are used to represent various constituent parts of the environment and edges are used to join the nodes together if there is a relationship between them. While this allows a user to visualise a complex environment in aggregate, it does not lend itself to visualising interaction between the nodes. To visualise interactions between nodes a completely different form of visualisation is usually employed in the form of a interaction diagram, also known as sequence diagram. In these types of diagrams, the predominant parameter is "time", i.e. a sequence of events. Often a whole dimension on the diagram, such as down the page or across the page, is reserved to represent this sequence. Furthermore, these diagrams are invariably for a single instance of interaction between the two nodes; they do not display multiple interaction instances together.A DG requires a diagram in two dimensions and interaction diagrams requires one dimension to be reserved for the flow of time, thus making them difficult to integrate directly together. Furthermore, even if you take the obvious approach to extend the diagram into the third dimension, where each edge is replaced with interaction diagrams in the z-plane, the resultant diagram is generally incomprehensible due to its size. This is because in a typical topology you have hundreds to thousands of edges and each edge will have hundreds or more interactions.This patent provides a method that allows the display of large topologies, as per DG methods, and adds the means to display in aggregate the interactions between the various nodes. This patent describes a method to represent interactions between two nodes in a compact manner which can be included in a DG representation of an application topology. The core of the idea is an animated glyph that can visually convey complex information about instances of an interaction. The visual construction of the glyph represents the source node, the number of outgoing and incoming communications and the status of those communications, and the animation of the glyph along edges between the nodes represents the pattern of the interaction between the two nodes, and the frequency of the specific pattern in relation to other patterns observed for the two nodes.This method is a new and novel way of representing transaction information with several benefits:- presents several attributes of interactional information in a simple diagram - allows interactional information to be displayed with an aggregate topology in a single DG, eliminating the need for the user to consult a second sequence diagram- allows the u...