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!View all the Fabric Data Days sessions on demand. View schedule
Hi,
Since yesterday there have been a massive increase in CU consumption in EventHouse UpTime operations. No changes have been done the Eventhouse setup. The data volume or quality haven't changed. The Eventhouse is running in Sweden Central.
Any ideas?
We ended up contacting MS support to investigate this further. It took a MS product manager to figure out, what has happened. Increased CPU consumption had triggered the upscaling. Now we know what to optimize.
I think it would be nice to have a more accurate documentation about the auto scaling logic.
Hi @Tröttöröö
Thanks for the update and for sharing those details. Glad to hear Microsoft Support helped identify the root cause.
Could you please confirm if the issue is now fully resolved after applying the optimizations ?
That's one of the reasons why we don't use auto scaling.
Hi @Tröttöröö
Just checking in to see if the previous response helped resolve your issue. If not, feel free to share your questions and we’ll be glad to assist.
We had similar issues and were told to reduce the retention period. Not sure if that is something you can consider.
Did this end up reducing CU consumption by auto scaling Eventhouse? If so, how immediate was the effect?
Not auto scaling. Within 24 hours, as expected.
Hello @Tröttöröö,
Here's some action you can do to check if it's normal :
Use the Capacity Metrics App
→ Check detailed CU usage per workload and timepoint.
Review Query Logs
→ Look for changes in query patterns or frequency.
Check EventHouse Settings
→ Confirm timepoint granularity, retention, and partitioning haven't changed.
Open a Microsoft Support Ticket
→ If no internal changes are found, Microsoft can investigate backend operations.
We can definitely see an increase that may need some more investigations !
Hope it can help you,
Antoine
Check out the November 2025 Fabric update to learn about new features.
Advance your Data & AI career with 50 days of live learning, contests, hands-on challenges, study groups & certifications and more!