Embed Performance - Diagnostic Steps

Hi Anonymous ,

According to your description, you said that the error is “6 gig memory limit was exceeded because of resource exceeded”,  and when this error occurs, you are performing a filtering operation on a very large dataset.

So the most possible reason causing the error may be your dataset is too large but your SKU is just only P1.

If you want to know the detail performance of your dataset, you can refer this docs: Monitor Power BI Embedded - Power BI | Microsoft Docs , it also provide you some guidelines about how to improve performance.

You can also consider purchase SKUs that support larger capacities.

Best regards,

Community Support Team Selina zhu

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly

The final solution was implemented by updating the dataset model to replace a many-to-many relationsip with an "Intersect", "Join" or some call it ia "Bridge" table containing a single inique key field that has a 1:n relationship betwen the two other tables.  The operation of the coss filtering works exactly the same but coss-filtering with RLS is now immediate and no longer runs out of memory.   The issue was not directly related to embedding, but appears to be related to RLS cross-filtering with many-many relationships.  Example:

Team User table had multiple teams for a user, multiple users on a team.   Needed to join the team to a list of accounts the team worked on.   Instead of joining the team between TeamUser.TeamKey and Account,TeamKey, added in the "Team" table that has only the unique  list of Team.TeamKey.   Join 1-many with Account and 1-many with TeamUser and the performance issue we were encountering was resolved.