Database size considerations¶
Since Rational BI loads the full relational database into the database embedded in the web browser, it’s important to consider performance and size when uploading data for reporting.
Databases are always loaded in full when requested which means that when requesting even a single row of data from a dataset, the system will transmit the associated SQLite database file to the client’s web browser.
The upside of this is that although the first access requires a full database transfer, any subsequent requests will be able to read the data directly from the local, now already loaded, database in the browser and access will be virtually free and instant.
Each database file has a unique UUID and is immutable which means that they can be efficiently cached and the web browser is typically able to cache accesses between report accesses. When data changes on the server, new database files will be made available and older files will fall out of browser caches.
Database files are transmitted in compressed, gzipped form and are typically very space efficient.
The maximum size of database files will vary with a number of factors such as the intended audience and other considerations. Larger files can safely be used for corporate users with known computing devices, but care should be taken to minimize the size of the database when publishing reports to broad audiences where minimum requirements can’t be guaranteed.
It is generally safe to use databases up to 25MB for most reporting targeted to unknown audiences. If the user base has typical reasonably current computing devices, database sizes up to 500MB or larger can be used.
When larger datasets than what can be loaded into the browser are required, consider using a remote dataset instead which queries data from a remote server instead of loading any data into the web browser.