from what an external consultant once told me, the cube cache is cleared once a partition is processed. This results in queries executed from these measures/partitions being slower. This is problematic since we daily process the partitions of the current year, which leads to the cache being cleared every morning.
This is what i could also observe and confirm, if a report was just run and than rerun immediatly after it is obviously running faster since the results has already been cached. My users also reported: over the day, times are going down since more and more data has been requested and is in cache already.
When i try to find stuff about caching in SSAS I stumble across Proactive Caching, which sounds exactly what i am looking for, but from what i am reading it is something totally different from what i am expecting. I do not care about changes being made in my relational database in the background.
What i am looking for is an option to somehow pre cache certain partitions after they have been freshly processed, so that once a user requests data from it, the cache is not totally empty and it must all be taken from "scratch".
Something like this seems to exist for SSRS
I also found something in this slideshare mentioning cube aggregations, which i should take a look into?
- This topic was modified 1 week, 1 day ago by ktflash.
I wanna be the very best
Like no one ever was