Analytical Query Processing

The GoodData Platform has been optimized to deliver high-performance query processing at scale. Multi-level caching ensures that response times are minimized as much as possible, and the logical data model provides an abstraction layer that ensures database independence. Depending on the requirements of your project, GoodData can deploy one of many databases as your datastore.

Key Benefit: The platform provides cube-style analytics processing without requiring users or developers to build, deploy, and manage analytics cubes.



Within the GoodData Platform, user requests are turned into a set of queries in a proprietary scripting language called MAQL DQL. Multi-dimension Analytical Query Language Data Query Language is a simplified database querying language designed specifically to interact with the GoodData datastore. This set of MAQL queries is passed through the logical data model to render the SQL queries required to interact with the physical data model, which returns the results to the requesting user.

Analytics Engine

Business users of the GoodData Portal can build visualizations to join information across multiple data sources. These queries are passed to the analytics engine, called the Extensible Analytics Engine, through REST APIs for processing. 

The Extensible Analytics Engine performs a number of query optimizations for processing and execution. Based on a series of mathematical operations, individual queries are partitioned horizontally by project and broken down into smaller, more efficient queries, the results of which are cached for future use. XAE inspects each calculation for sub-queries existing in caches across the entire multi-tenant platform, optimizing for performance and freshness of results.

  • This Advanced Query Optimization utilizes an internal variant called MAQL Algebra to systematize query execution.
  • The platform relies heavily on function shipping for optimal query performance and scaling under concurrent load.

Key Benefit: The platform can perform multi-dimensional analysis against relational data without relying on predefined aggregates or cubes. This flexibility is supported by data query technologies including columnar datastores, projections, and clustering.

Powered by Atlassian Confluence and Scroll Viewport.