Forum Discussion
Does semantic model update skip unchanged files?
I have quite a large set of data that I use to feed a PBI report and I am thinking about ways to make it update faster. One Idea that came up was creating a second database with historical data that will not be edited anymore apart from the one with recent data, which gets edited daily.
If the semantic model update skips unchanged files, this could - in theory - increase the update speed, does anyone know if this works?
Have you looked into Incremetal refresh?
This lets you:
Indicate how far back to refreshProvide an update field and only items with a new date will refresh
Using Premium have a live connection for some data, refresh import and finally stored non updating for the oldest.
Official page:
https://learn.microsoft.com/en-us/power-bi/connect-data/incremental-refresh-overview
Video guide (Guy in a Cube also made some great ones!):
https://www.youtube.com/watch?v=kH_mcS9Bv9E
2 Replies
- SamWiseOwl
Super User
Have you looked into Incremetal refresh?
This lets you:
Indicate how far back to refreshProvide an update field and only items with a new date will refresh
Using Premium have a live connection for some data, refresh import and finally stored non updating for the oldest.
Official page:
https://learn.microsoft.com/en-us/power-bi/connect-data/incremental-refresh-overview
Video guide (Guy in a Cube also made some great ones!):
https://www.youtube.com/watch?v=kH_mcS9Bv9E - danextian
Have you looked into using incremental refresh on new records or records that have changed? If this isn't feasible for some reason, I'd split the database into at least two - one for the historical data and the other one for current - and load them as separate tables to be connected using relationships. You can disable the refresh on historical tables. Note: they need to be loaded separately and not be combined in one query as that will still pull the data from the historical table upon refresh.
