Use Query Labels in Microsoft Fabric Warehouse for Better Workload Monitoring

As organisations adopt Microsoft Fabric Data Warehouse for analytics, reporting, and ETL pipelines, thousands of queries can run daily across different workloads—Warehouses, BI reports, data pipelines, ad-hoc analysis, and automated processes.

Without clear identifiers, it becomes difficult to determine:

• Which queries power critical reports and dashboards

• Which pipelines are consuming the most compute

• Which workloads are causing performance bottlenecks

Query Labels in Fabric Warehouse provide a simple but powerful way to solve this problem. By tagging SQL queries with meaningful labels, teams can easily monitor, analyse, and troubleshoot query workloads using Fabric’s monitoring tools.

What Are Query Labels?

A query label is metadata that can be attached to a SQL query in Fabric Warehouse or the SQL analytics endpoint.

The label acts as an identifier that allows engineers to:

• Track queries in Query Insights views

• Monitor workload performance

• Analyse execution history

• Troubleshoot slow or expensive queries

Instead of searching raw SQL text, you can simply filter queries by their label name.

Query Label Syntax

Adding a label to a query is simple. You include the OPTION (LABEL = '<label>') clause at the end of the SQL statement.

Syntax

SELECT ...FROM ...OPTION (LABEL = 'label_name');

The label can contain any meaningful identifier describing the query workload.

Example 1: Label a Dashboard Query

If a query powers a sales dashboard, you can tag it with a label.

SELECT

FinanceKey,      

DateKey,     

OrganizationKey,      

DepartmentGroupKey,      

SUM(Amount) AS TotalAmount

FROM dbo.FactFinance

WHERE OrganizationKey = 13AND DepartmentGroupKey = 103

GROUP BY FinanceKey, DateKey, OrganizationKey, DepartmentGroupKey

OPTION (LABEL = 'SALES_DASHBOARD');

This makes it easy to track the query’s performance across executions.

Example 2: Monitor Long Running Queries

Once labelled queries run, they can be monitored using Query Insights system views:

SELECT * FROM queryinsights.long_running_queries

WHERE last_run_command LIKE '%SALES_DASHBOARD%'ORDER BY median_total_elapsed_time_ms DESC;

This helps identify slow-running queries associated with specific workloads.

Example 3: View Query Execution History

You can also analyse execution history for labelled queries:

SELECT *FROM queryinsights.exec_requests_history

WHERE label = 'SALES_DASHBOARD'ORDER BY submit_time DESC;

This view allows engineers to monitor:

• Execution frequency

• Runtime performance

• Historical workload trends

Why Query Labels Matter

1. Better Workload Visibility

Query labels make it easy to identify which workloads are running in the warehouse, such as dashboards, reports, or ETL jobs.

2. Faster Performance Troubleshooting

When a performance issue occurs, labels help quickly identify the responsible workload.

Examples:

ETL_PATIENT_LOAD

POWERBI_WAITLIST_DASHBOARD

FINANCE_MONTHLY_REPORT

3. Improved Governance

In enterprise environments with multiple teams, labels help organise and monitor workloads across departments.

Typical workload categories include:

• Dashboards

• ETL pipelines

• Reporting queries

• Ad-hoc analytics

4. Monitoring Critical Queries

Labels allow teams to track high-value queries powering:

• Production dashboards

• Data pipelines

• Regulatory reporting workloads

Practical Tips for Using Query Labels

1.Use a Naming Convention

Consistent naming helps maintain clarity across teams.

Example:

DASHBOARD_<name>ETL_<process>REPORT_<department>PIPELINE_<workflow>

Examples:

• ETL_PATIENT_ADMISSIONS_LOAD

• DASHBOARD_WAITLIST_METRICS

• REPORT_FINANCE_MONTHLY

2.Label Important Workloads

Focus on labelling queries that are:

• Part of production ETL pipelines

• Used in dashboards or reports

• Known to consume significant compute

Long running or performance sensitive

3.Combine With Monitoring Views

Query labels are most useful when used with Query Insights views, including:

• queryinsights.exec_requests_history

• queryinsights.long_running_queries

These views provide detailed telemetry about query execution.

Example: Healthcare Data Warehouse

In a healthcare analytics platform (such as NHS reporting), labels could track different workloads across the medallion architecture.

Label	Purpose
ETL_PATIENT_ACTIVITY_LOAD	Bronze → Silver ingestion pipeline
ETL_WAITLIST_AGGREGATION	Gold metric aggregation
DASHBOARD_AE_PERFORMANCE	Emergency department dashboard
REPORT_FINANCE_MONTHLY	Finance reporting

This allows platform teams to quickly identify which workloads affect performance.

Key Considerations

• Query labels must be manually added to SQL statements

• Query insights views may take up to 15 minutes to show completed queries

• Governance relies on maintaining consistent naming standards

Despite these considerations, query labels significantly improve the observability and management of warehouse workloads.

Conclusion

Query labels are a small feature with a big impact in Microsoft Fabric Data Warehouse. By simply tagging SQL queries with meaningful identifiers, teams gain better visibility into workloads, faster troubleshooting capabilities, and improved operational governance.

In large data platforms supporting dashboards, pipelines, and enterprise reporting, query labels help turn complex query activity into manageable and observable workloads.

Call to Action

If you are building solutions in Microsoft Fabric, consider adopting query labels as part of your development standards.

Start by:

1. Identifying critical queries powering reports, dashboards or pipelines

2. Adding labels using OPTION (LABEL = 'name')

3. Monitoring labelled queries using Query Insights views

4. Establishing naming conventions across engineering teams

This simple practice can greatly improve performance monitoring, workload management, and governance in your Fabric Warehouse environment.

References

Microsoft Documentation:

Query labels - Microsoft Fabric | Microsoft Learn