Forum Discussion
Table relationship causes severe performance drop
KayceVC -
Regarding the snowflaking - There are 3 tables that could be removed and incorporated into their respective dimension tables:
Months -> Dates (can create a M-to-M relationship with Agreement Revenue)
Staff Teams -> Staff
Customer Teams -> Companies
you have circular relationships betwen tables which is causing poor performance. tryy to avoid it
Sorry gooranga1 , I need to slow down and read closer. I did change this to a single direction and it made no noticable difference to the report. Creating, deleting, saving measures and columns still takes between 13 - 15 minutes.
what is the actual pbix file size when saved?
14,270KB
A couple of optimization thoughts:
1. Star schema is preferred to snowflake. At query time, less relationships = better performance.
2. Remove any unnecessary columns. This will reduce memory consumption of the model. The most memory-intensive columns are those with many distinct values.
Thank you for the advice, Natel, but these were some of the first things I did in optimizing this report. Unfortunately, the link to this date table is still causing a problem for me.