Announcing billing for Workspace monitoring
Workspace Monitoring, currently in preview is an observability feature within Fabric that enables monitoring capabilities across two key areas:
- Logging: gain insights into system events and actions.
- Performance metrics: get insights on performance metrics and queries in Fabric.
Workspace monitoring allows Fabric developers and admins to access detailed logs and performance metrics for their workspaces. This helps troubleshoot performance issues, investigate errors, optimize queries, and minimize data downtime.
Since the preview announcement, we have added support for Mirrored database operation logs in Workspace monitoring. These logs can be used to monitor execution and performance of your mirrored database, including data replication, table changes, mirroring status, failures, and replication latency for mirrored databases and tables.
We have also introduced template reports for accelerated insights into query performance and optimizing operations. You can easily download the Power BI template or Real-time Dashboard templates from fabric-toolbox for configuration over the Monitoring Eventhouse.
We are announcing activation of billing for this feature that will commence on March 10, 2025, please read on for more details on billing.
Architecture overview
Workspace Monitoring operates by leveraging two Fabric workloads:
- Monitoring Eventstream: handles real-time event brokerage and messaging. Used to push logs and metrics from Fabric workloads to the monitoring Eventhouse.
- Monitoring Eventhouse: manages structured event data storage and querying.
These two workloads are automatically created once you enable monitoring within your workspace. The workloads are managed by the system and are read-only for customers.
Activation of billing for Workspace monitoring
Billing for the underlying components such as Eventstream and Eventhouse for Workspace monitoring will commence on or after March 10, 2025. Since the Eventstream and Eventhouse are native items in Fabric, it utilizes Fabric capacity units like other Fabric workloads. There are no additional charges beyond the Capacity Unit (CU) consumption listed below.
Listed are the consumption charges by operation type; please refer to the dedicated pricing pages for each workload for more details:
- Eventstream per hour: 0.22 CU hour.
- Eventstream data traffic per GB: 0.342 CU hour.
- Eventhouse UpTime compute costs: 1 CU hour/active v-core.
- Eventhouse storage billing: storage is billed separately from the Fabric capacity units. Data ingested into the Monitoring Eventhouse is charged across OneLake cache storage ($0.246 per GB per month) and OneLake Standard Storage ($0.023 per GB per month).
This pricing structure ensures alignment with Fabric’s consumption-based model, making monitoring costs predictable and scalable as usage grows.
Note: there are no charges for the Eventstream processor or Eventstream connector; only the flat charge and data traffic charges apply.
Monitor usage in Metrics App
You can monitor the Fabric capacity unit consumption and storage related charges for workspace monitoring in the Fabric capacity metrics app using Item name ‘Monitoring Eventhouse’ and ‘Monitoring Eventstream’. The following image shows a sample compute page from monitoring capacity in the Fabric Capacity Metric app:
Operation name: EventHouse UpTime
Operation name: Eventstream per hour and data traffic per GB
Monitor Workspaces with monitoring enabled
Capacity admins can easily view the workspaces that have enabled workspace monitoring.
What’s next?
As we continue to enhance Workspace monitoring capabilities, we are focused on onboarding more Fabric items, expanding the feature set and optimizing architecture for improved performance and cost efficiency. We also plan on adding ability to write logs from multiple workspaces to a single capacity.
Stay tuned for more updates as we continue to unlock powerful observability capabilities for your data workloads in Fabric.
Have questions or feedback? Drop them in the comments or reach out to us directly.
