New Capacity health using Microsoft's official measures
Each capacity page now shows the same indicators as the Capacity Metrics App: average utilization over the last hour, the last 24 hours and the last 7 days, rejection status and accumulated overload over the last 60 minutes. The numbers are updated every 5 minutes.
New Plain-language capacity diagnosis
A Diagnosis card classifies the capacity as Healthy, At risk, Throttling or Rejecting and explains why, without jargon. It shows whether the capacity is right-sized (average versus peak, bursts), how long users wait (P95), time conflicts between user access and refreshes (with the items worth rescheduling) and refreshes that got slower.
New A CU% chart that explains the spikes
With Show events turned on, the period's refreshes and jobs appear as dots on the CU% curve, with failures in red. You can see at a glance which run caused each spike. The chart also gained a time range filter (1h, 6h, 24h, 3 days and 7 days).
New Capacity page with tabs
The capacity detail is now organized into Overview, Throttling (with a counter) and Workspaces. In the workspace list, a Git badge shows which ones are connected to source control.
New Daily consumption in CU-seconds
The Capacity Dashboard now shows today's and yesterday's consumption (with the change), the success rate over the last 24 hours, failed refreshes and a bar chart of the last 30 days with your SKU's limit line. The calculation uses data collected by FabricBoard itself.
New Executive KPIs on the home page
The Dashboard gained the KPIs management asks for: unique users and views (14 and 30 days), active items, refresh failure rate over the last 7 days and average CU utilization over 24 hours, each linking to its detail screen. Trend charts show tenant growth over 90 days (workspaces, artifacts and active users).
New Usage & Adoption screen
A new Governance screen with the most viewed content (top 15), items with no usage in the period (candidates for archiving) and day-by-hour heatmaps, in your organization's time zone, for views and refreshes. It's the foundation for deciding what to keep, reschedule or turn off.
New Refresh reliability
A new tab in Data Refreshes with a 14-day board per model (one square per day), success rate and average duration, worst first. The Trend column compares last week's duration with the week before and highlights what got worse.
New Refresh schedule with time-slot collisions
A new Schedule tab in Data Refreshes shows what is scheduled and points out the time slots where several refreshes compete for the capacity at the same time.
New Tenant Settings history
The Governance Radar gained a Tenant Settings tab. FabricBoard records the Fabric admin settings every day and shows on a timeline what changed (enabled/disabled and allowed groups).
New Risky sharing on the Radar
Two new cards on the Governance Radar. One flags any use of Publish to web in the last 90 days as critical. The other warns about content shared with the entire organization.
New Workspace matrix in the Inventory
A new tab with a heatmap crossing capacity, workspace and item type, with totals for workspaces, items and access entries, plus a breakdown of access by principal type. It's a quick view of where your tenant's content is concentrated.
Improved Contextual help on metrics screens
An (i) icon next to cards, charts and columns explains what each number means. The capacity page, the Capacity Dashboard, Reliability and Usage & Adoption also gained screen-level help.
Improved More accurate throttling
The throttling alert now considers only the last hour, so it resolves itself once the capacity is back to normal. The screen still shows 7 days of history, with pagination, and makes clear what is happening now and what is history.
Improved More compact Capacity Dashboard
Smaller cards and shorter charts, so you see more without scrolling.
Fixed Notebooks, Lakehouses, Pipelines and SQL endpoints in the Inventory
Some Fabric items weren't imported because Microsoft returns them in a different format from Power BI items. Now any item type that comes back in the scan is added to the inventory.
Fixed Clipped CU% charts
Capacity charts capped the axis at around 12% and hid peaks above that. The scale now follows the actual values.
Fixed Direct links and page reload
Opening a screen from its direct address or pressing F5 could show a 404 error, and the direct link to the Capacity Dashboard opened the home page. Both issues are fixed.