Activity¶
Two dashboards analyse the Power BI activity log: User Activity shows who is active and how, System Load shows when the tenant is busy.
Both read the activity events MetaMiner collects from the Power BI activity log, summarised by the usage rollup job. They follow the common filters (range, environments, capacities); see Command Center. Hours are UTC.
User Activity¶
Open it with Select Analysis > User Activity.
Image of the User Activity dashboard: five KPI cards, Daily Active Users bar chart, Activity Type Breakdown donut, Hourly Activity Distribution, and the Top Power Users table
../images/dashboards/user-activity.png
KPI cards¶
| Card | What it shows |
|---|---|
| Active Identities | Distinct identities (user IDs) with any activity in the range |
| Peak Daily Users | The busiest single day: distinct identities that day |
| Report Views | ViewReport events in scope |
| Total Events | All operations in scope |
| Avg Events / Identity | Total events divided by active identities, over the whole range |
Panels¶
| Panel | What it shows | Click |
|---|---|---|
| Daily Active Users | Unique identities with any activity, per day | A day opens Day Activity (operations that day) |
| Activity Type Breakdown | Share of events by category: View, Refresh, Export, Scanner / Admin API, Other | - |
| Hourly Activity Distribution | Average events per hour of day | An hour opens Hour Of Day (the days behind that hour) |
| Top Power Users | The 15 identities with the most events. Columns User ID, Type, Total Events, Report Views, Refreshes, Exports, Active Days, Profile. Click a column header to sort. | A total opens User Activity By Day for that identity |
| Activity Type Mix | Share of events by category, with the share of view operations | - |
| Daily User Engagement | The seven days with the most identities, plus Other days | - |
| User Engagement Summary | Distinct identities, peak single-day users, total report views and total events | - |
Type classifies each identity:
| Type | Rule |
|---|---|
| MetaMiner SP | The service principal registered for a MetaMiner environment (MetaMiner's own collection traffic) |
| System | Blank, PowerBI or None identities |
| Service | Other GUID identities, app identities, and accounts named svc-, robotic or adm- |
| Member | All remaining user principal names |
Profile labels each top user: MM Scanner, System, Automated, Heavy Viewer (more than 50 report views), Heavy Refresh (more than 50 refreshes), Exporter (more than 20 exports) or Mixed.
System Load¶
Open it with Select Analysis > System Load.
Image of the System Load dashboard: Peak Hour, Business-Hours Load, Off-Hours Load and Quiet Trough cards, the Weekly Activity Heatmap with the Absolute/Normalized toggle, and the daily activity calendar
../images/dashboards/system-load.png
KPI cards¶
| Card | What it shows |
|---|---|
| Peak Hour | The busiest hour of day, with its events |
| Business-Hours Load | Share of events between 08:00 and 18:00 UTC |
| Off-Hours Load | Share of events outside 08:00-18:00 UTC |
| Quiet Trough | The quietest hour of day, with a workload label: Business-hours workload (70 percent or more in business hours), Mixed / 24x7 (45-70 percent) or Off-hours / batch-heavy |
Panels¶
| Panel | What it shows | Click |
|---|---|---|
| Weekly Activity Heatmap | Weekday x hour of day. Each cell is the average events per active day of that weekday. Absolute shows raw averages; Normalized scales each weekday row to its own peak. Below it, a daily activity calendar of the range. | A cell opens Heatmap Cell Days; a calendar day opens Day Activity |
| Weekday vs Weekend | Average events per active day for weekdays and weekend days (raw totals would mislead, because a range has few weekend days) | - |
| Hour-of-Day Distribution | Total events per hour of day across the range | An hour opens Hour Of Day |
| Daily Breakdown | One row per day: Date, Events, Users, Report Views, Dashboard Views, Refreshes, Other, View %, with a period total row. View % is labelled human-driven (25 percent or more), mixed (5-25 percent) or machine/batch. | A row opens Day Activity |
Distinct counts do not add up
The Users column counts distinct identities per day. The period total is its own distinct count, not the sum of the column.
Questions these pages answer¶
- How many people, service accounts and automated processes use the tenant, and how intensively?
- Who are the heaviest viewers, refreshers and exporters?
- How much of the traffic is MetaMiner's own collection or other automation?
- When is the tenant busiest, and is the load business-hours or batch-driven? Useful for refresh scheduling and capacity planning.
Settings¶
These pages have no thresholds of their own. The maximum range is web.apps.pbiusage.max.range.days (default 92). See Server Properties.