Performance optimization and near self-experience are lucrative features that customers are always looking for. Reusing data sources across the application, dynamic slicer and dice facility are available to a primitive extent in SAP Lumira Designer. Data utility features in VBX expands this further and provides advance runtime data source manipulation capabilities while minimizing the number of instances of the data source in the dashboard. Let’s take a deep dive into how VBX’s data utility can help you overcome performance bottlenecks.
Reutilization of data sources
In Lumira designer there is a data selection feature for components that allows the developer to restrict/filter the data to be shown during the run time. This feature provides of reusing the same data source without creating multiple copies. Unfortunately, data selection works only on measures, not dimensions (However users will be able to select individual members of a dimension).
There is a Data Utility feature in all the VBX Charts which allows for dimension & measure selections, change in the order of dimensions, measure aggregation, suppression type. This feature is available as a discrete component called Data Utility Source under the custom data source category. When creating huge dashboards, if similar data utility rules are applied in various parts of the dashboard, it might be tedious for the users to recreate the same set of rules for many individual components. Data utility sources can be used here to store these sets of rules. Creating a data utility source is akin to creating virtual copies of the “Source” data source and selecting the required dimensions and measures that should appear in the charts. These data sources can now be assigned to any component like a typical data source.
As shown in the below image we are using 2 components that are assigned to 2 virtual copies of our data source and the main data source (DS_4) is assigned to the data utility component.
Once the data source is been assigned to the data utility component, data selection can be done for the virtual copies separately including dimensions. This really reduces the run time and shows only the information required in the initial view.
Data Utility provides an edge over its native counterpart by allowing the users to select dimensions also.
Near Self-service experience
Allowing the user to change the granularity of the data set and to select the way in which the data is presented in runtime mimics a near real-time experience to the user.
Navigation Panel in Lumira designer allows the user to change displayed dimensions, their order, and measures in the data source.
However, the change impacts the data source and all the visualizations bound to this data source are affected. But when using data utility data sources, we are using a virtual data source for each chart and the dimensions and measures in all these virtual data sources can be controlled from a single cockpit component called – “Data Utility” which can be added to the layout of the application just like the navigation panel. This allows the users to change the data utility rules defined in the virtual data sources during runtime.
As seen in the above GIF, two virtual copies (DS_REGION & DS_STORE) have been created and assigned to a chart and table each. Here you can observe that the rendering of the component happens in the user-defined layout. It provides the Initial View appearance having Dimensions and Measures in the run time with columns and rows appearing as buttons “C” and “R” where C represents the Columns and R represents the Rows. You can now only reassign the Dimensions to the Rows or Columns by a single click on “R” or “C” at a time. Also, you cannot reassign the Measures to the Rows or Columns as they remain configured based on the design time. You can only remove or add the Measures and at least there should be one measure being added. There is an option that you can deselect the members of both the Dimensions and Measures at any time.
The dashboard created allows the user to change the granularity on the fly to get more insights. This can be extended to the users to select the data source from connections in run time using the back-end connection technical component.
Let us consider two scenarios where we need to view the key figures across various dimensions in four different visualizations with and without the Data Utility Component.
As shown in the above image, the profiling statistics shows that the data source time on startup is 5.3 seconds.
With Data Utility:
Using the Data Utility component reduces half the data source time (2.4 Seconds) on startup.
Information provided in this document is intended to be used only as a learning aid. While Visual BI Solutions has used reasonable care in compiling this document it does not warrant that this information is error-free. Visual BI Solutions assumes no liability whatsoever for any damages incurred by usage of this documentation.
This is intended to be a living document with changes to be made over time.
The steps outlined below are only for learning purposes. Try this out in your own environment at your own risk. Information is provided ‘as-is’, without warranty of any kind. Visual BI will not be liable for any damages or consequences thereof.