A. Anatomy of Scheduled Refresh

In Power BI Service, a dataset (semantic model) connected to Dataverse can be refreshed on a schedule.
Set frequency, time zone, and time slots. Refresh pulls the latest records via the connector without manual action.

B. Basic Prerequisites

• Power BI Pro (per user) or Premium capacity (per capacity).
• Dataverse connector used in Power BI Desktop.
• Publish the PBIX to a workspace where you can manage refresh.
• Correct permissions in Dataverse (read tables used).

C. Cloud Connections vs Gateway

Dataverse is cloud-native; no on-prem gateway needed. For on-prem sources (SQL/Oracle/etc.),
install a Standard (enterprise) or Personal gateway. You can mix: Dataverse (cloud) + on-prem SQL via gateway.

D. Data Source Credentials

Use OAuth2 for Dataverse. Credentials are encrypted in Power BI Service.
Update via Workspace → Semantic model → Settings → Data source credentials.
If a token expires or permissions change, edit and re-auth here.

E. Enterprise Gateway (Standard Mode)

• Install on a server (always on, stable network).
• Shared across users/data sources; centrally administered.
• Supports high-throughput and load balancing with clusters.
• Use for production; avoid personal gateways for teams.

F. Personal Gateway

Tied to one user; installs on their PC; not shareable. Good for individual refresh needs.
PC must be online at refresh time. For teams/production, move to Standard gateway.

G. Gateway vs Cloud: When to Use What

No gateway for Dataverse/SharePoint Online/other cloud sources.
Gateway for on-prem SQL/Oracle/Files/SAP etc.
Hybrid models use both; configure each source appropriately in the model.

H. Sensitivity Labels & Governance

• Apply MIP labels (e.g., Confidential) to reports/datasets.
• Labels persist to exports, supporting DLP and auditing.
• Combine with workspace roles and Dataverse security roles.

I. Parameters (Dev/Test/Prod)

Create parameters for Environment URL or optional filters (e.g., last N months).
After publishing, set parameter values per workspace via the Service, then rebind data sources if needed.

J. Query Caching (Premium)

Premium capacities can cache query results to accelerate dashboards. It doesn’t replace refresh;
it speeds up repeated queries while the scheduled refresh keeps the data current.

K. Refresh Types (Manual, Scheduled, Incremental)

• Manual: On demand in Desktop or Service.
• Scheduled: Automated via Service (time + frequency).
• Incremental: Refreshes only a window of data; define RangeStart/RangeEnd.

L. Monitoring & Refresh History

In the Service, open Semantic model → Settings → Refresh history to see status, duration, and errors.
Use this to tune schedules and quickly fix failures.

M. Common Data Sources (CE + Beyond)

• Dataverse (CE): Cloud connector; no gateway.
• SQL Server on-prem: Needs gateway + credentials.
• Files on-prem (CSV/Excel): Needs gateway.
• SharePoint Online: Cloud connector; no gateway.

N. Security Model Alignment

Align workspace roles (Viewer/Contributor/Member/Admin) with Dataverse security roles.
Use RLS (row-level security) when needed; ensure filters align with CE business units/teams.

O. Optimizing Model & Queries

• Limit columns/rows; remove unused fields.
• Push filters to the source (Dataverse query folding).
• Use star schema; avoid many-to-many where possible.
• Leverage incremental refresh for large CE tables.

P. Premium vs Pro Considerations

Premium adds higher refresh frequency, larger model sizes, query caching, and advanced governance.
Pro is fine for smaller CE footprints; upgrade as data volume and concurrency grow.

Q. Query Diagnostics & Logs

Use Power Query diagnostics in Desktop to find slow steps. In the Service, check refresh error details.
For gateways, review gateway logs (performance counters, network latency).

R. Refresh Scheduling Strategy

• Pick off-peak windows to reduce CE/API pressure.
• Stagger multiple datasets to avoid concurrency spikes.
• Use different frequencies by business criticality.

S. Semantic Model Refresh: Local vs Cloud

• Local PBIX: Refresh in Desktop (uses your credentials), then Publish.
• Cloud: Scheduled refresh runs in Service using stored credentials/gateway.
• Keep schema stable; changes may require re-binding sources/parameters.

T. Troubleshooting Playbook

• “Unable to connect”: verify credentials & permissions in CE.
• Gateway offline: restart service; check updates/firewall/proxy.
• Timeouts: reduce data, fold queries, incremental refresh.
• 504/429: throttle; reduce frequency; stagger datasets.

U. Usage Metrics & Alerts

Enable Usage metrics reports. With Power Automate, email/Teams-notify owners on refresh failures.
Consider Azure Monitor for deeper operational insights in Premium.

V. Versioning & Deployment

Store PBIX/M queries in source control (e.g., Git, Power Query text export).
Use deployment pipelines (Dev→Test→Prod) and parameterize environment URLs.

W. Workspace Design

• Separate workspaces by environment or domain.
• Least-privilege roles; keep owners minimal.
• Document refresh windows in the workspace description.

X. eXceptions & Limits to Know

• Dataset size and refresh frequency limits vary by Pro/Premium.
• API throttling on Dataverse; design efficient queries.
• Some transformations may break query folding → slower refresh.

Y. Your End-User Experience

Communicate refresh times so users know when data is “fresh”.
Pin key visuals to dashboards; enable subscriptions for stakeholders after refresh completes.

Z. Zero-to-One Setup (Quick Checklist)

1. Build the report in Desktop using Dataverse connector.
2. Define Parameters (Environment URL, ranges) and enable Incremental Refresh where needed.
3. Publish to the correct workspace.
4. Set data source credentials (OAuth2) in the Service.
5. For any on-prem sources, install & map Standard Gateway.
6. Configure scheduled refresh (time zone, slots, frequency).
7. Apply sensitivity labels and review RLS.
8. Monitor refresh history; add alerts/Power Automate notifications.