Forum Discussion
Super UserYour Fabric compute session disconnected because the notebook session completed or timed out.
Ok checked and fixed:
with an FTL4 Capacity, you need to create a Spark Pool with the configurations below in the Workspace Settings → Data Engineering/Science section:
Then you need to create an Environment Artifact and add this new Spark Pool to the Environment Artifact:
Publish the changes and then in a notebook you need to use this new Environment Artifact that has this Small Spark Pool which fits in the capacity of FTL4:
Best,
Jacek
Hi jaryszek ,
Thank you for providing the error details. The error message suggests that the notebook session ended because the Spark request could not be accommodated by the available capacity at that moment. Specifically, the TooManyRequestsForCapacity error indicates that the Spark compute or API limits were reached when the job was submitted.
Even with an F64 Trial capacity, Spark workloads are subject to concurrency and capacity limitations. If other notebooks, Spark jobs, pipelines, or background workloads are running simultaneously on the same capacity, new requests may be throttled until resources are freed. This can result in the notebook session disconnecting and showing a session completed status.
To investigate further, please review the Monitoring Hub and Job Management section in Workspace Settings at the time of the issue. These tools can help determine if there were active or queued Spark workloads using capacity. Additionally, it would be beneficial to know if the notebook was executed interactively or via a pipeline, if multiple users share the trial capacity, and whether this behavior is consistent or isolated.
The following documentation may assist in understanding Spark concurrency, capacity limits, and workload monitoring in Fabric:
Concurrency limits and queueing in Apache Spark for Fabric - Microsoft Fabric | Microsoft Learn
Job concurrency and queue monitoring (Preview) - Microsoft Fabric | Microsoft Learn
Job queueing in Apache Spark for Fabric - Microsoft Fabric | Microsoft Learn
If you are able to provide further information, I can conduct a more detailed review to help identify the root cause.
Thank you.