Circular Dependency Error

[koralka]

Hey all,
I’m dealing with a very unusual issue in a mixed-mode (Import and Direct Query) Power BI semantic model (over 5GB, used across many dependent reports) - I am not the author of that model.
Just a note as well that there were no recent changes made to that model, the error just appeared. Ihad a look at different possible solutions/causes but I am not sure how to approach this...  I'll try to summarise everything. 

There is a calculated table (_targetRevenueTable) in Import mode that contains several calculated columns, which suddenly started showing a circular dependency error, details below:

Data source error: A circular dependency was detected: _targetRevenueTable[_targetRevenueTable-0ccd4068-558e-4373-a3b2-3eabdf58ccc3], 21a49dea-b623-429b-9a57-f4397cc811f5, _targetRevenueTable[date], _targetRevenueTable[_targetRevenueTable-0ccd4068-558e-4373-a3b2-3eabdf58ccc3]

• The table itself is Import and not linked to any DirectQuery tables.

• All source tables used in the calculated columns (website_job, _dimDate, etc.) are also Import.

• Power BI Desktop does not flag any syntax errors in the DAX for the table or for any of the calculated columns.

• The table view looks normal, no red underlines, and the model designer shows the table as Import.

• The same error has happened once before but disappeared on its own without any model changes.

• A partition is showing for this table in Power BI Desktop, but it’s not visible in Tabular Editor, which makes the setup feel inconsistent or potentially corrupted.

• The above partition is what the error references: [_targetRevenueTable-0ccd4068-558e-4373-a3b2-3eabdf58ccc3]....

The model is very large and widely used, so it's difficult to evaluate downstream impact... This column is used in multiple measures and other calculated columns across many reports - I had a look in Tabular Editor and DAX Studio.

Because the model is shared by many teams and feeds many datasets, any structural changes (e.g., replacing calculated columns with measures or recreating the table) are difficult to assess for impact. 

I'm trying to find best practice for safely repairing this without breaking dependent reports.

I wouold really appreciate some help.

Thank you 

[tayloramy]

Hi @koralka, 

THe first thing I would do is make a copy of the model in a dev workspace that has no downstream dependencies, so you can do whatever you need to to get it refreshing again without impacting anything, and then once you know how to fix it, you can implement the changes in the production model. 

I would start by deleting and re-creating that table. Those partitions seems sus to me. 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution.

[koralka]

Hi @tayloramy 
Thank you... Yes, I am making changes in dev. The issue is - I start to think that the error message isn't actually pointing to the actual error. I've 'fixed' (made some changes) so what could have possibly be causing the circular dependency error on those targetRevenue tables and refresh failed with another circular dependency error on completely different tables. Interestingly, both are calculated tables from other tables that pull from SharePoint

[tayloramy]

Hi @koralka, 

Interesting. Could you share more details? Could we see a sdcreenshot of the model view so we can see relationships? 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution.

[v-aatheeque]

Hi @koralka 

Following up to confirm if the earlier responses addressed your query. If not, please share your questions and we’ll assist further.

[v-aatheeque]

Hi @koralka 

Have you had a chance to look through the responses shared earlier? If anything is still unclear, we’ll be happy to provide additional support.

[koralka]

Hello, thank you for your reply. Apologies for delayed response, I was on annual leave. 

I wouldn't be able to provide screenshots of the model as it's company's data but also, this model in large and relationships aren't really clear from just looking at the visual represenation of them. I managed to solve the issue with duplicating some tables and changing the relationships between them. I am still unsure how it's possbile that it was working fine and no wors was done on those tables for months and then it suddenly broke...