I have a Model that has three steps DB - Scheme - Model and loads Incrementally. This is a big model, with over than 100M records in the main Fact table.
The process is being managed by the QMC Tasks and runs once an hour.
Occasionally, there is no new data from the DB and I want to be able to skip the reload of the Scheme + Model for performance reasons.
Options as I see it now:
Exit Script - When using QMC always saves the qvw, and an wild empty model appears
Partial Reload - Haven't worked it out completely, should I dig in further?
In case you have a high volume and low frequent source that you do not want to reload everytime due to a presence of low volume and high frequent source in your UI app then you can use Partial reload to run the script that loads from low volume and high frequent source and will skip running the script for high volume data.
However from QMC point of view you might need to create two sets of tasks.
One for Full reload low frequent and another one for Partial reload high frequent like every 5 min or so.
I think you need one layer more: DB - QVD Generator - Scheme - Model in your data-architecture.
Another rather dirty way would be not to exit the script else to produce intentionally an error which breaks the execution and keeps your application on the previous version.
