This time we’re going bigger than ever. Fabric, Power BI, SQL, AI and more. We're covering it all. You won't want to miss it.
Learn moreDid you hear? There's a new SQL AI Developer certification (DP-800). Start preparing now and be one of the first to get certified. Register now
Hi
We have a massive data set. We enabled incremental load on one transactional table. and published to service. In service, the dataset failed due to this issue
| The database was evicted and the operation cancelled to load balance the CPU load on the node |
Solved! Go to Solution.
Individual partitions cannot exceed 10GB. A refresh request cannot exceed 5 hours. You may be hitting one or both of these limits just with this semantic model but not with the others.
Use bootstrapping for the partition definitions and then manually refresh the historical / cold partitions vial the XMLA endpoint.
Thanks for the article. I have other models also established in the same way. I dont experience this error in other incremntal load configured model. So it looks weird that this is happening only in this model alone
Individual partitions cannot exceed 10GB. A refresh request cannot exceed 5 hours. You may be hitting one or both of these limits just with this semantic model but not with the others.
Check out the April 2026 Power BI update to learn about new features.
Sign up to receive a private message when registration opens and key events begin.
If you have recently started exploring Fabric, we'd love to hear how it's going. Your feedback can help with product improvements.
| User | Count |
|---|---|
| 10 | |
| 9 | |
| 8 | |
| 8 | |
| 7 |
| User | Count |
|---|---|
| 50 | |
| 27 | |
| 23 | |
| 23 | |
| 22 |