Currently I have 2 QV models which show different context but are from the same logic flow in the system.
the first one is reloaded 1-2 times a day and the other is refreshed every 15-30 minutes.
the source data is completely different for each model except one meta data table.
I was approached by our product team and was asked to check if there is a possibility to merge this models even though the data is almost entirely different.
the problem is the reload intervals, and my question is of two parts:
1) How do I separate the the script of this model in a way that part of it is fully reloaded and part is partially loaded.
2) I know there is a partial reload method (using ADD and REPLACE in load statements) in 'File' but how is it handled in the console?
will be glad to get suggestions regarding this topic.
I don't know so much about partial reload, but I would suggest something different:
- You say only this one table is present in both models, yes? How big is that? Will there be any added value in loading this only once (I assume that the metadata does not change often in a day)?
-> If there is, I suggest you try building a small additional app loading only that one data table and then loading that via a BINARY LOAD just as often as you need it. => Remember a BINARY has to be the VERY FIRST command in the script (even before a hidden script if there should be one). <=> That should work, but depending on how your model looks, there might be complications if there are any APPLYMAP lines in that LOAD - since the mapping_tables cannot be loaded beforehand. It has to fit into your data_models.
