Forum Discussion
rsQueryTimeoutExceeded
- Anonymous9 months ago
Hi jesushmmadrid ,
Using a Composite Model can definitely help in separating the date logic while still keeping both fact tables consistently filterable, and applying group by/ summarize patterns in key measures can also reduce the size of the filter context and improve performance.
Please follow these articles for reference :
Optimization guide for Power BI - Power BI | Microsoft LearnData reduction techniques for Import modeling - Power BI | Microsoft Learn
Data reduction techniques for Import modeling - Power BI | Microsoft Learn
Regards
Hi everyone,
Thank you very much for all your suggestions. I have tested the recommended solutions, including using two separate date tables and replacing USERELATIONSHIP with TREATAS, but the issue still persists and I continue hitting resource limits.
What I have tried:
1. Creating two separate date tables (AdmissionDate and InvoiceDate).
This improved performance, but it created a different problem. Several report pages combine metrics from both HOSMA_Negocio and HOSMA_LineasFactura using the same date slicers. With two date tables, slicers only filter one fact table unless I introduce bidirectional relationships or bridge tables, which causes ambiguity and performance issues. As a result, numbers between business and invoicing no longer match.
2. Replacing USERELATIONSHIP with TREATAS.
This reduces the load, but I still get timeouts when filtering large date ranges. I believe the indirect path through IdEpisodio_TipoEA forces very large filter contexts.
3. Simplifying the model and reducing columns.
Even with these optimizations, heavy visuals still fail.
What I need clarification on:
The core challenge seems to be that both fact tables are connected through IdEpisodio_TipoEA and are used together on multiple report pages. I need both fact tables to be filtered consistently, but invoicing measures must use a different date context, without breaking existing pages or duplicating the model entirely.
Before restructuring the whole model, I would like to ask:
Is there any recommended pattern for scenarios where two fact tables are connected, both need consistent filtering in some pages, but a subset of measures must use a different date path?
Would a calculation group, a disconnected date table with TREATAS, or a composite model pattern be more appropriate?
I appreciate any further guidance. It is a complex scenario and I would like to avoid breaking existing report logic.
Thank you again for your help.
Hi jesushmmadrid ,
Using a Composite Model can definitely help in separating the date logic while still keeping both fact tables consistently filterable, and applying group by/ summarize patterns in key measures can also reduce the size of the filter context and improve performance.
Please follow these articles for reference :
Optimization guide for Power BI - Power BI | Microsoft Learn
Data reduction techniques for Import modeling - Power BI | Microsoft Learn
Data reduction techniques for Import modeling - Power BI | Microsoft Learn
Regards