Overview of the Latest Features and Improvements in Jet Analytics
Jet Analytics 2019 (version 19.2 of the Jet Data Manager) introduces a number of important new features and improvements. Below is an introduction to the changes made in this version of Jet Analytics.
As we continue to develop the Jet Data Manager, we add new options and features that improve performance.
For instance, the 18.10 release included significant performance improvements for conditional lookup fields
In the new release, we have included a feature to help you take advantage of these improvements with only a few clicks. With performance recommendations, you are presented with a list of possible optimizations and a list of the affected objects. You can deselect any optimizations you do not want to apply, but otherwise, you just need to click OK and deploy your project to improve execution times.
While we could just enable the options and features that improve performance automatically, we take a cautious approach to rolling out the new features. New projects will default to using the improved settings, but we will not change settings in existing projects.
Improved Performance for Incremental Load and History Tables
One of the major advantages of using the Jet Data Manager is that improved performance is just a deployment away when we optimize the code generated by the platform. This release includes performance tweaks to incremental load and history. Especially for tables with history enabled, the gains are substantial. In our tests, history-enabled tables execute about twice as fast as before - or even faster. However, we cannot guarantee performance improvements on that level in all cases since it very much depends on the complexity of the table.
Data Lineage for Custom Views
In the new release, you can map the ‘input’ and ‘output’ fields in a custom view to enable tracing all the way from data source to semantic layer. This information is used both in the visualization options in the software as well as in the documentation it generates.
In previous versions of the Jet Data Manager, a custom view would create fields ‘out of the blue’, since the Jet Data Manager cannot automatically map fields from data sources to the fields created by a custom view.
Shared Semantic Layer
Dynamic Row-Level Security
In the new release of the Jet Data Manager, you can set up row-level security on models in the shared semantic model based on the contents of a data warehouse table. With this feature you can have a single role that grants access to all domain users and then differentiate what the individual user has access to by supplying a table the contains information about the user and data relationship.
Add Descriptions for use in SSAS Tabular
You can now add descriptions of tables, fields and measures in a semantic model and deploy them to SSAS Tabular endpoints. The descriptions are useful for helping end users understand the data in Power BI.
Support for Perspectives on SSAS Tabular
Perspectives can improve the user experience of a complex model for the end-user by hiding the tables, fields and measures the user does not need. You can now add perspectives to semantic models and deploy them to SSAS Tabular endpoints.
Due to the improvements to incremental load and history logic on tables in the data warehouse, some settings on tables has been deprecated:
Enable BK hash key (Table Settings -> Performance tab): Enabling history on the table will create a similar hashed version of the business key, should you need it.
Target-based incremental load (Table Settings -> Data Extraction tab): Enabling history on the table will give you the same benefits with identical performance.
Use left outer join (Table Settings -> History tab): This option had very limited use for troubleshooting, but with the new logic, there is no longer any use for it.