Error when trying to build the Semantic Model for Power BI

[pfekrati]

🐛 Problem

After adding historical data the semantic model fails to build with the following error:

Data source error: Expression.Error: Query execution lacks memory resources to complete (80DA0007): Partial query failure: Low memory condition (E_LOW_MEMORY_CONDITION). (message: 'bad allocation (E_LOW_MEMORY_CONDITION)', details: '').
[0]Kusto.Data.Exceptions.KustoServicePartialQueryFailureLowMemoryConditionException: Query execution lacks memory resources to complete (80DA0007): Partial query failure: Low memory condition (E_LOW_MEMORY_CONDITION). (message: 'bad allocation (E_LOW_MEMORY_CONDITION)', details: '').
Timestamp=2026-01-29T21:27:41.9276181Z
ClientRequestId=KPBI;da28ac00-f5fc-4ba7-9df5-2ecb44a9b174;3138df23-7f3e-4884-88bf-4413de5280c2;2ebdf4a5-c6ca-4efb-a9ac-775f33162a27
ActivityId=e934732f-a425-4efd-a1c0-2560369e3f8a
ActivityType=GW.Http.CallContext
ServiceAlias=KUTRIDENTHOSTEREUS2
MachineName=KEngine00000T
ProcessName=Kusto.WinSvc.Svc
ProcessId=3200
ThreadId=10740
ActivityStack=(Activity stack: CRID=KPBI;da28ac00-f5fc-4ba7-9df5-2ecb44a9b174;3138df23-7f3e-4884-88bf-4413de5280c2;2ebdf4a5-c6ca-4efb-a9ac-775f33162a27 ARID=e934732f-a425-4efd-a1c0-2560369e3f8a > GW.Http.CallContext/e934732f-a425-4efd-a1c0-2560369e3f8a)
MonitoredActivityContext=(ActivityType=GW.Http.CallContext, Timestamp=2026-01-29T21:27:25.0945493Z, ParentActivityId=e934732f-a425-4efd-a1c0-2560369e3f8a, TimeSinceStarted=16833.0738 [ms])ErrorCode=
ErrorReason=
ErrorMessage=
DataSource=
DatabaseName=
ClientRequestId=
ActivityId=00000000-0000-0000-0000-000000000000
UnderlyingErrorCode=80DA0007
UnderlyingErrorMessage=Partial query failure: Low memory condition (E_LOW_MEMORY_CONDITION). (message: 'bad allocation (E_LOW_MEMORY_CONDITION)', details: '')

. . The exception was raised by the IDbCommand interface.
Cluster URI: WABI-US-EAST2-C-PRIMARY-redirect.analysis.windows.net
Activity ID: fba2f2a1-2c56-4e2d-b59b-08d1b32b2421
Request ID: 20039106-b14f-8d31-9dd2-f0cc7eb9cd55
Time: 2026-01-29 21:27:42Z

👣 Repro steps

TODO: Add repro steps below:

1. Deploy FinOps Hubs in Fabric
2. Export historical data for ~ 1 year (13GB worth of data)
3. Build the Semantic Model
4. See error

🤔 Expected

Semantic model should builds successfully after ingesting the historical data

[microsoft-github-policy-service]

Thank you for submitting this issue! We try to respond within 3 business days. In the meantime:

• 📖 Check the troubleshooting guide for common issues
• 💬 Ask questions in Discussions
• 📅 Join our biweekly office hours for live help from the team

[RolandKrummenacher]

Hi @pfekrati

Can you please confirm whether this behavior is related to the v13 release of the FinOps Toolkit, or if you were already observing this before upgrading?

This will help us understand if this might be tied to recent changes.

[flanakin]

@pfekrati A few thoughts:

• I don't think the semantic model matters because we're using DirectQuery.
• What report/page are you seeing this on?
• Do you see this when you make a direct call to the Hub database (e.g., Costs | limit 100)?
• What Fabric capacity do you have configured?
• Do you have the minimum consumption on the eventhouse set? I doubt it, but asking just in case. To learn more minimum consumption, see Minimum consumption.

Since this is actually an issue with Fabric, you can file a support request with them to have them dig in and give you more explicit insights into what the issue might be. This will go to the Real-Time Intelligence / Azure Data Explorer team.