Data Access Using InMemory MOLAP

Foresight Analytics Platform implements the following methods for working with data:

ROLAP. Get data by means of SQL queries from various DBMS including relational and columnar in real time. An SQL query to DBMS is executed in real time each time data is accessed. The execution result is sent to the BI server and converted into a data matrix that is stored in BI server RAM. Handling of SQL query execution result may take a long time, particularly if the volume of requested data is large. This approach is characterized by its flexibility, as it allows for loading only the data that is currently required. At the same time, with large data volumes, not only can data retrieval latencies increase significantly, but also memory consumption, which grows proportionally to the volume of extracted data. article.
InMemory MOLAP. Data retrieval from a specialized InMemory MOLAP server included in Foresight Analytics Platform. InMemory MOLAP server ensures quick retrieval and analysis of multidimensional data and delivers data without significant delays for data source opening or aggregation. Data sources are standard cubes and cube views ad virtual cubes that are built based on them.

InMemory MOLAP

InMemory MOLAP is a technology that ensures data loading from sources to RAM with later data access "on the go" without having to access the hard disk or the DBMS, which significantly speeds up data access and processing.

To store multidimensional data outside DBMS and quickly process it using the InMemory MOLAP technology, Foresight Analytics Platform implements the following:

File cache. Data is stored in a proprietary internal file format and displayed fully or partially in RAM as needed. The memory-mapped representation not only ensures data persistence but also enables the implementation of higher-performance data processing algorithms. File cache supports additional data processing methods. For details about file cache see the Working with File Cache article.
Global (cross-session) cache. Data is stored in a standard in-memory matrix format, with persistence achieved through saving and restoring data as BLOB within a relational DBMS. For details about global cache see the Working with Global Cache article.

Users can apply a specific caching mechanism depending on the tasks being solved.

InMemory MOLAP is implemented on the InMemory MOLAP server. For details about installing and general settings of InMemory MOLAP server see the Installing InMemory MOLAP Server article.