Join us at FabCon Atlanta from March 16 - 20, 2026, for the ultimate Fabric, Power BI, AI and SQL community-led event. Save $200 with code FABCOMM.
Register now!Calling all Data Engineers! Fabric Data Engineer (Exam DP-700) live sessions are back! Starting October 16th. Sign up.
Hey all, seems I'm like the only person to experience this but I have a big hairy report, it has so many tables and connections in it I've almost lost count.
It refreshes perfectly fine in the online service, takes about 8 minutes to fully refresh each morning. However, if I have to work with it I can't do a full refresh in desktop without it failing, and when it fails it usually sites and "error in step : merge columns in....", however if I open that table and step through it it's perfectly fine, no errors. Also, the error messages are different each time.
The only way for me to refresh in the desktop is to manually refresh tables one-by-one until it just works again and does a full refresh. It's very annoying.
Usually, I need to refresh the big FACT tables manually, and then the rest will work.
Here's a couple of points :
- Semantic model includes SQL tables, sharepoint excel sheets and Analysis Cube connections
- Refreshes perfectly fine online everyday
- I have to manually refresh single tables until it will finally allow a full auto refresh all
- Usually, after I have manually refreshed the FACT tables it will work
- The FACT tables are mostly all Analysis Cube connections (maybe this is why???)
It's really annoying and takes up so much time to have to manually refresh until it allows a full refresh. Help?!
Hi @Moehimby
Thank you for reaching out to the Microsoft Fabric Forum Community.
@samratpbi @danextian @DaleT Thanks for the inputs.
I hope the information provided by users was helpful. If you still have questions, please don't hesitate to reach out to the community.
Thanks for the welcome. Unfortunately there hasn't been any tips or info that help me resolve the issue yet.
I have one semantic model that does that. Before refreshing in desktop, I make sure that the preview has finished loading for queries involved.
Hi, I also face similar issues many times with large data models with many tables, relationships and reference tables and many Power Query transformations. I think when we try to do full refresh desktop tries to validate everything which consumes quite a few memories. I also noticed, recently this has increased compared to a few years back. Now, either due to my laptop is getting older or in recent Power BI versions, some additional things has been introduced.
Interesting I didn't think of this as my laptop is about 10 years old too. Funny thing is tho - I can still pull up web pages or write emails when it does refresh so it doesn't act like the machine is out of memory.
Hi,
Can you share the full error message? Did you do a lot operations in Power Query? For example, join two or more tables, look up value from another table.
Error messages vary but are usually some variation of "Table Sales failed at step : Merge Columns 3", then all the other tables will show "Cannot refresh due to dependency on table Sales".
Yes, there are probably 20 tables in the model, most lookups are done via merging in power query, not dax.
The interactions among tables could be the cause. I have a small model in which I do a lot of Power Query operations. Usually it fails the first refresh, but it succeeds when I try it again. It's hard to debug because it doesn't fail again after the first refresh. I would suggest pushing all the Power Query operations to the backend like MySQL or SQL Server.
Join the Fabric FabCon Global Hackathon—running virtually through Nov 3. Open to all skill levels. $10,000 in prizes!
Check out the September 2025 Power BI update to learn about new features.