fabric-paginated-report-sch

Fabric Paginated Report Batch Executor

A production-ready framework for Microsoft Fabric and Power BI paginated report automation. This Python-based batch executor enables automated generation of paginated reports with dynamic parameters from multiple data sources including Lakehouse, Semantic Models, Warehouses, and JSON arrays. Built for enterprise data engineering workflows with features like retry logic, OneLake storage integration, and Azure pipeline orchestration via REST API.

Features

✅ Four Flexible Parameter Sources

Semantic Model (Power BI Dataset) - Query data behind reports with DAX
Lakehouse (Delta Lake) - Native Spark SQL integration
JSON Array - Simple, direct input for testing
Warehouse (SQL Server) - T-SQL queries for enterprise scenarios

✅ Output Options

OneLake storage with date-based folder hierarchy
Multiple export formats (PDF, Excel, Word, PowerPoint, PNG)
Automatic file archival and organization

✅ Enterprise-Ready

Retry logic with exponential backoff for failed reports
Continue on failure (processes all parameters even if some fail)
Managed identity authentication
Configurable timeouts and retry intervals
Pipeline integration with detailed logging

✅ User-Friendly

No code changes needed to update parameters
Business users can maintain parameter lists
Comprehensive logging with multiple levels
Detailed error messages with context

Quick Start
Architecture
Parameter Sources
Setup Instructions
Usage Examples
Pipeline Configuration
Advanced Features
Troubleshooting
Best Practices

Quick Start

1. Upload the Notebook

Open your Fabric workspace
Create a new notebook
Import paginated_report_batch_executor.ipynb
Attach to a Lakehouse (if using Lakehouse parameter source)

2. Choose Your Parameter Source

Select one of the four parameter sources based on your needs:

Source	Best For	Setup Effort
Semantic Model	Report-driven parameters, business users	Low
Lakehouse	Large lists, data engineers	Low
JSON	Testing, static lists	Minimal
Warehouse	Enterprise, complex SQL	Medium

3. Run a Test

Create a pipeline using pipeline/pipeline_definition.json
Configure parameters (see Usage Examples)
Run manually to test
Verify files in OneLake (Files/reports/archive folder)
Enable scheduled trigger once working

Architecture

Pipeline (Scheduled/Manual Trigger)
  │
  ├─ Parameters: report_id, static_params, report_partitioning_source, etc.
  │
  ▼
Fabric Notebook: Report Batch Executor
  │
  ├─ Cell 1-2: Parameters & Imports
  │
  ├─ Cell 3: Input Validation & ParameterLoader
  │   ├─ InputValidator class
  │   └─ ParameterLoader class (supports 4 sources)
  │
  ├─ Cell 4: Token Manager & Validation
  │   ├─ TokenManager class (automatic refresh)
  │   └─ Parse & validate all parameters
  │
  ├─ Cell 5: Power BI API Functions
  │   ├─ handle_api_response()
  │   ├─ initiate_report_export()
  │   ├─ poll_export_status()
  │   └─ download_report_file()
  │
  ├─ Cell 6: File Storage Functions
  │   ├─ sanitize_filename()
  │   ├─ generate_file_name()
  │   └─ save_to_onelake()
  │
  ├─ Cell 7: Main Execution Function
  │   └─ execute_report_with_retry()
  │
  ├─ Cell 8: Load Special Parameter Values
  │   └─ ParameterLoader.load()
  │       ├─ Semantic Model (DAX query via sempy)
  │       ├─ Lakehouse (Spark SQL)
  │       ├─ JSON (Direct parsing)
  │       └─ Warehouse (pyodbc + T-SQL)
  │
  ├─ Cell 9: Main Execution Loop
  │   └─ FOR EACH partitioning parameter value:
  │       ├─ Merge static + special params
  │       ├─ Execute report (Power BI API)
  │       ├─ Poll & download
  │       ├─ Save to OneLake
  │       └─ Retry 3x on failure
  │
  └─ Cell 10: Results & Exit
      └─ Return summary to pipeline

Execution Flow

Parameter Loading: Load partitioning parameter values from configured source
Parameter Merging: Combine static parameters with each special value
Report Generation: Execute paginated report via Power BI REST API
File Storage: Save to OneLake with date-based folder structure
Error Handling: Retry failures with exponential backoff, continue to next parameter
Summary: Report total successes, failures, and file locations

Parameter Sources

1. Semantic Model (Recommended for Business Users)

Query data from Power BI semantic models using DAX.

Advantages:

Uses trusted data from existing reports
Row-Level Security automatically enforced
Includes business logic and calculations
Fast with semantic layer caching

Configuration:

{
  "report_partitioning_source": "semantic_model",
  "semantic_model_workspace_id": "workspace-guid",
  "semantic_model_dataset_id": "dataset-guid",
  "semantic_model_dax_query": "EVALUATE FILTER(DISTINCT('DimCustomer'[CustomerName]), 'DimCustomer'[IsActive] = TRUE)"
}

Example DAX Queries:

-- Get all active customers
EVALUATE FILTER(
    DISTINCT('DimCustomer'[CustomerName]),
    'DimCustomer'[IsActive] = TRUE
)

-- Get top 10 customers by sales
EVALUATE TOPN(
    10,
    SUMMARIZECOLUMNS(
        'DimCustomer'[CustomerName],
        "TotalSales", SUM('FactSales'[SalesAmount])
    ),
    [TotalSales], DESC
)

-- Get customers with sales in last 12 months
EVALUATE CALCULATETABLE(
    DISTINCT('DimCustomer'[CustomerName]),
    DATESINPERIOD('DimDate'[Date], TODAY(), -12, MONTH)
)

See setup/sample_semantic_model_queries.dax for more examples.

2. Lakehouse (Recommended for Data Engineers)

Query Delta tables in Lakehouse using Spark SQL.

Advantages:

Native Spark SQL (no connection strings needed)
Best performance for large lists (1000+)
Easy maintenance via Lakehouse UI
ACID compliance and version control
Zero authentication overhead

Setup:

-- Run in Lakehouse notebook
CREATE TABLE parameter_config (
    Category STRING,
    ParameterValue STRING,
    IsActive BOOLEAN,
    SortOrder INT
) USING DELTA;

-- Insert sample data
INSERT INTO parameter_config VALUES
    ('MonthlyReportCustomers', 'Acme Corp', true, 1),
    ('MonthlyReportCustomers', 'TechStart Inc', true, 2),
    ('MonthlyReportCustomers', 'Global Solutions', true, 3);

Configuration:

{
  "report_partitioning_source": "lakehouse",
  "lakehouse_table": "parameter_config",
  "lakehouse_category": "MonthlyReportCustomers",
  "lakehouse_column": "ParameterValue",
  "lakehouse_filter": ""
}

See setup/create_lakehouse_parameter_table.sql for complete schema.

3. JSON Array (Recommended for Testing)

Provide parameters directly as JSON array.

Advantages:

Simplest setup (no infrastructure)
Perfect for testing
Good for static, small lists
Version control friendly

Configuration:

{
  "report_partitioning_source": "json",
  "report_partitioning_values": "[\"Acme Corp\", \"TechStart Inc\", \"Global Solutions\"]"
}

Use Cases:

Testing and development
Proof of concept
One-time or ad-hoc reports
Very small lists (< 10 items)

4. Warehouse (Recommended for Enterprise)

Query Warehouse tables using T-SQL.

Advantages:

Familiar T-SQL syntax
Complex SQL logic (joins, CTEs, stored procedures)
Row-Level Security (RLS)
Column-Level Security (CLS)
Integration with SQL Server workflows

Setup:

-- Run in Warehouse
CREATE TABLE dbo.ParameterConfig (
    Category NVARCHAR(100),
    ParameterValue NVARCHAR(500),
    IsActive BIT,
    SortOrder INT
);

-- Insert data
INSERT INTO dbo.ParameterConfig VALUES
    ('MonthlyReportCustomers', 'Acme Corp', 1, 1),
    ('MonthlyReportCustomers', 'TechStart Inc', 1, 2);

Configuration:

{
  "report_partitioning_source": "warehouse",
  "warehouse_name": "EnterpriseWarehouse",
  "warehouse_table": "dbo.ParameterConfig",
  "warehouse_column": "ParameterValue",
  "warehouse_category": "MonthlyReportCustomers"
}

Setup Instructions

Prerequisites

Microsoft Fabric workspace
Paginated report published to workspace
Lakehouse for OneLake storage (automatically created with workspace)

Step 1: Import Notebook

Navigate to your Fabric workspace
Click New → Import notebook
Select paginated_report_batch_executor.ipynb
If using Lakehouse source, attach notebook to Lakehouse

Step 2: Set Up Parameter Source

Choose one of the four options:

Option A: Semantic Model

Identify Power BI semantic model with parameter data
Note workspace GUID and dataset GUID
Write DAX query (see setup/sample_semantic_model_queries.dax)
Test query in Power BI Desktop

Option B: Lakehouse

Attach notebook to Lakehouse
Run setup/create_lakehouse_parameter_table.sql
Run setup/sample_data.sql for test data
Verify: SELECT * FROM parameter_config

Option C: JSON

Create JSON array with values
Validate JSON syntax
Ready to use!

Option D: Warehouse

Create Warehouse in workspace
Run table creation SQL (see config examples)
Insert parameter data
Grant read permissions

Step 3: Create Pipeline

Go to Pipelines in Fabric
Create new pipeline
Add Notebook activity
Select paginated_report_batch_executor notebook
Configure parameters (see Usage Examples)
Add triggers if needed (Daily, Weekly, Monthly)

Step 4: Test

Run pipeline manually
Monitor notebook execution
Verify files in OneLake (Files/reports/archive folder)
Check execution logs for any errors
Enable scheduled trigger once working

Usage Examples

Example 1: Monthly Customer Reports (Semantic Model)

Scenario: Generate monthly sales report for each active customer using data from Power BI semantic model.

Pipeline Parameters:

{
  "report_id": "12345678-1234-1234-1234-123456789abc",
  "workspace_id": "workspace-guid",
  "output_format": "PDF",
  "static_params": "{\"start_date\": \"2024-01-01\", \"end_date\": \"2024-12-31\"}",
  "report_partitioning_column": "Customer",
  "report_partitioning_source": "semantic_model",
  "semantic_model_workspace_id": "workspace-guid",
  "semantic_model_dataset_id": "dataset-guid",
  "semantic_model_dax_query": "EVALUATE FILTER(DISTINCT('DimCustomer'[CustomerName]), 'DimCustomer'[IsActive] = TRUE)",
  "archive_to_onelake": "true",
  "max_retries": "3"
}

Result: PDF reports saved to OneLake at Files/reports/archive/2025/01/30/, one per customer.

Example 2: Regional Reports (Lakehouse)

Scenario: Generate quarterly report for each region using parameter list from Lakehouse.

Setup Lakehouse:

INSERT INTO parameter_config (Category, ParameterValue, IsActive, SortOrder)
VALUES
    ('QuarterlyRegions', 'North America', true, 1),
    ('QuarterlyRegions', 'Europe', true, 2),
    ('QuarterlyRegions', 'Asia Pacific', true, 3),
    ('QuarterlyRegions', 'Latin America', true, 4);

Pipeline Parameters:

{
  "report_id": "report-guid",
  "workspace_id": "workspace-guid",
  "output_format": "XLSX",
  "static_params": "{\"quarter\": \"Q1\", \"year\": \"2024\"}",
  "report_partitioning_column": "Region",
  "report_partitioning_source": "lakehouse",
  "lakehouse_table": "parameter_config",
  "lakehouse_category": "QuarterlyRegions",
  "lakehouse_column": "ParameterValue"
}

Result: Excel reports for 4 regions saved to OneLake.

Example 3: Testing with JSON

Scenario: Test report generation for 3 specific customers before rolling out to all.

Pipeline Parameters:

{
  "report_id": "report-guid",
  "workspace_id": "workspace-guid",
  "output_format": "PDF",
  "static_params": "{\"start_date\": \"2024-01-01\", \"end_date\": \"2024-01-31\"}",
  "report_partitioning_column": "Customer",
  "report_partitioning_source": "json",
  "report_partitioning_values": "[\"Test Customer A\", \"Test Customer B\", \"Test Customer C\"]"
}

Result: 3 test reports generated quickly without setting up database tables, saved to OneLake.

Example 4: Enterprise Warehouse with RLS

Scenario: Generate reports for customers visible to current user based on Row-Level Security.

Warehouse Setup:

-- Create table with RLS
CREATE TABLE dbo.CustomerReporting (
    CustomerName NVARCHAR(200),
    AssignedTo NVARCHAR(100),
    IsActive BIT
);

-- Create RLS policy
CREATE FUNCTION dbo.fn_CustomerSecurityPredicate(@AssignedTo NVARCHAR(100))
RETURNS TABLE WITH SCHEMABINDING AS
RETURN SELECT 1 AS Result
WHERE @AssignedTo = USER_NAME() OR USER_NAME() IN (SELECT UserName FROM dbo.Admins);

CREATE SECURITY POLICY dbo.CustomerReportingPolicy
ADD FILTER PREDICATE dbo.fn_CustomerSecurityPredicate(AssignedTo)
ON dbo.CustomerReporting WITH (STATE = ON);

Pipeline Parameters:

{
  "report_partitioning_source": "warehouse",
  "warehouse_name": "EnterpriseWarehouse",
  "warehouse_table": "dbo.CustomerReporting",
  "warehouse_column": "CustomerName",
  "warehouse_category": ""
}

Result: Each user only generates reports for their assigned customers (RLS enforced).

Pipeline Configuration

Required Parameters

Parameter	Description	Example
`workspace_id`	Fabric workspace GUID	`"12345678-1234-..."`
`report_id`	Paginated report GUID	`"87654321-4321-..."`
`output_format`	Export format	`"PDF"`, `"XLSX"`, `"DOCX"`
`static_params`	Fixed parameters (JSON)	`"{\"start_date\": \"2024-01-01\"}"`
`report_partitioning_column`	Parameter to loop through	`"Customer"`
`report_partitioning_source`	Parameter source type	`"semantic_model"`, `"lakehouse"`, `"json"`, `"warehouse"`

Source-Specific Parameters

Semantic Model:

semantic_model_workspace_id
semantic_model_dataset_id
semantic_model_dax_query

Lakehouse:

lakehouse_table
lakehouse_category
lakehouse_column
lakehouse_filter (optional)

JSON:

report_partitioning_values

Warehouse:

warehouse_name
warehouse_table
warehouse_column
warehouse_category

Optional Parameters

Parameter	Description	Default
`archive_to_onelake`	Save to OneLake	`"true"`
`max_retries`	Retry attempts per report	`"3"`
`export_timeout_seconds`	Max seconds to wait for export	`"600"`
`poll_interval_seconds`	Seconds between status polls	`"5"`

Advanced Features

1. Multiple Parameter Loops

For complex scenarios with multiple loop parameters:

Option A: Use Semantic Model with SUMMARIZECOLUMNS

EVALUATE
SUMMARIZECOLUMNS(
    'DimGeography'[Region],
    'DimProduct'[Category]
)

This returns all combinations of Region × Category. The notebook will loop through each row.

Option B: Pre-compute combinations in Lakehouse

INSERT INTO parameter_config (Category, ParameterValue, IsActive, SortOrder)
SELECT
    'RegionCategoryCombo',
    CONCAT(Region, '|', Category),
    true,
    ROW_NUMBER() OVER (ORDER BY Region, Category)
FROM region_category_combinations;

Then parse the combined value in your report.

2. Conditional Parameter Lists

Filter parameters based on date, status, or other criteria:

Semantic Model (DAX):

EVALUATE
FILTER(
    DISTINCT('DimCustomer'[CustomerName]),
    'DimCustomer'[IsActive] = TRUE
    && CALCULATE(SUM('FactSales'[Amount]), DATESINPERIOD('DimDate'[Date], TODAY(), -6, MONTH)) > 10000
)

Lakehouse (SQL):

{
  "lakehouse_filter": "ValidFrom <= CURRENT_TIMESTAMP() AND (ValidTo IS NULL OR ValidTo >= CURRENT_TIMESTAMP())"
}

3. Dynamic File Naming

Files are automatically named with this pattern:

Report_{SpecialParamName}_{SanitizedValue}_{Timestamp}.{format}

Example:

Report_Customer_AcmeCorp_20250130_143022.pdf

Special characters are sanitized for filesystem compatibility.

4. Error Handling and Retry

Retry Logic:

3 retry attempts per report (configurable)
Exponential backoff: 30s, 60s, 120s
Continues to next parameter on failure

Error Categories:

Power BI API errors (authentication, report not found, etc.)
Export failures (timeout, invalid parameters, etc.)
File storage errors (OneLake write failures, etc.)

All errors are logged with detailed messages.

5. Monitoring and Notifications

Console Logging: Real-time output shows:

Parameter loading progress
Each report execution (5 steps per report)
Success/failure status
File paths and sizes
Execution summary

Pipeline Notifications: Configure webhook URL for notifications:

Success: Includes file count, total size, paths
Failure: Includes error details

Example Webhook Payload (Success):

{
  "status": "success",
  "pipelineRunId": "run-guid",
  "successCount": 5,
  "failCount": 0,
  "totalSize": "12.5 MB",
  "files": [
    "/reports/2025/01/30/Report_Customer_AcmeCorp_20250130_143022.pdf",
    "/reports/2025/01/30/Report_Customer_TechStart_20250130_143035.pdf"
  ]
}

6. Performance Tuning Parameters

Fine-tune the notebook’s performance and behavior for specific scenarios:

Parameter	Default	Range	Description
`download_chunk_size_mb`	1	1-100	Download chunk size in MB for large files
`file_size_warning_mb`	500	10-5000	File size threshold for warnings
`connection_timeout_seconds`	30	5-300	API connection timeout
`download_timeout_seconds`	120	30-600	File download timeout
`param_loader_retry_attempts`	3	1-10	Parameter loading retry count
`param_loader_retry_delay_seconds`	5	1-60	Delay between parameter loading retries
`token_refresh_interval_minutes`	45	5-55	Auto token refresh interval

When to Adjust:

Large Files (Reports > 100MB):

{
  "download_chunk_size_mb": "5",
  "file_size_warning_mb": "1000",
  "download_timeout_seconds": "300"
}

Slow Network Connections:

{
  "connection_timeout_seconds": "60",
  "download_timeout_seconds": "240"
}

Long-Running Batches (1000+ reports):

{
  "token_refresh_interval_minutes": "30"
}

Note: Tokens expire after 60 minutes. Refresh interval should be less than 55 minutes.

Unreliable Parameter Source:

{
  "param_loader_retry_attempts": "5",
  "param_loader_retry_delay_seconds": "10"
}

Token Refresh Mechanism:

Tokens are automatically refreshed based on token_refresh_interval_minutes
Prevents authentication failures during multi-hour batch executions
No user intervention required
Logged with timestamp for audit purposes

Troubleshooting

Common Issues

1. No Parameter Values Loaded

Symptoms: “No parameter values loaded! Check your configuration.”

Solutions:

Semantic Model: Verify DAX query returns data in Power BI Desktop
Lakehouse: Check table exists and Category matches exactly (case-sensitive)
JSON: Validate JSON syntax at jsonlint.com
Warehouse: Verify table exists and connection works

2. Authentication Errors

Symptoms: 401 Unauthorized, token errors

Solutions:

Verify managed identity is enabled for workspace
Check service principal has access to:
- Report workspace (Contributor or higher)
- Semantic model (Read permission)
- Lakehouse (Read permission)
- Warehouse (SELECT permission)
Refresh Azure AD token if expired

3. Report Export Timeout

Symptoms: “Export timeout after 600 seconds”

Solutions:

Simplify report (reduce data, remove complex visuals)
Increase timeout in notebook (currently 10 minutes)
Check report parameters are valid
Verify underlying data sources are accessible

4. Slow Performance

Symptoms: Reports take very long to generate

Solutions:

Semantic Model: Ensure query is optimized, use DISTINCT instead of VALUES
Lakehouse: Add indexes on Category and IsActive columns
Report: Optimize report design, reduce data volume
Parallel Processing: Consider splitting large batches across multiple pipelines

Debug Mode

To run notebook interactively for debugging:

Open notebook in Fabric
Set default values in Cell 1 for all parameters
Run cells one by one
Check output of each cell
Fix issues and re-run

Best Practices

1. Parameter Source Selection

Scenario	Recommended Source
Business users maintain list	Semantic Model or Lakehouse
Need Row-Level Security	Semantic Model or Warehouse
Large lists (1000+)	Lakehouse
Testing/Development	JSON
Complex SQL logic	Warehouse
Highest performance	Lakehouse

2. File Organization

OneLake Structure:

Files/
  /reports/
    /archive/
      /{YYYY}/{MM}/{DD}/
        Report_Customer_Value_Timestamp.pdf

Example:

Files/reports/archive/2025/01/30/Report_Customer_AcmeCorp_20250130_143022.pdf

3. Scheduling

Daily Reports: Run at off-peak hours (e.g., 6:00 AM)
Weekly Reports: Run Monday morning
Monthly Reports: Run on 1st or last day of month
Avoid: Running multiple large batches simultaneously

4. Security

✅ Always use managed identity for authentication
✅ Grant minimal required permissions to workspaces and data sources
✅ Use Row-Level Security (RLS) when available in semantic models
✅ Audit parameter access and report generation
✅ Validate and sanitize all input parameters
✅ Use proper error handling to avoid exposing sensitive information
❌ Never hardcode credentials or connection strings
❌ Never commit secrets or GUIDs to version control

5. Maintenance

Regularly review and clean up inactive parameters
Monitor pipeline success rates and execution times
Archive or delete old reports from OneLake to manage storage
Update DAX queries when semantic model schema changes
Test after any Fabric platform updates
Document parameter meanings and categories for business users
Review and optimize timeout and retry settings based on actual performance

Support and Contributing

Getting Help

Check Troubleshooting section
Review example configurations in config/ folder
Check Fabric platform status
Review notebook execution logs
Contact your Fabric admin or support team

File Structure

/
├── paginated_report_batch_executor.ipynb  # Main notebook
├── README.md                               # This file
├── setup/
│   ├── create_lakehouse_parameter_table.sql
│   ├── sample_data.sql
│   └── sample_semantic_model_queries.dax
├── config/
│   ├── example_semantic_model.json
│   ├── example_lakehouse_mode.json
│   ├── example_json_mode.json
│   └── example_warehouse_mode.json
└── pipeline/
    └── pipeline_definition.json

Version History

v1.0 (2025-11-03)

Production-ready paginated report batch execution framework
Four flexible parameter sources: Semantic Model (DAX), Lakehouse (Spark SQL), JSON, Warehouse (T-SQL)
Automatic token refresh for long-running batches (handles 1-hour token expiration)
OneLake archival with date-based folder hierarchy
Retry logic with exponential backoff for robust error handling
Configurable performance tuning parameters for enterprise scenarios
Comprehensive input validation and SQL injection protection
Pipeline integration with detailed logging and progress tracking
Continue-on-failure support for large batch operations
Managed identity authentication

License

This project is provided as-is for use within Microsoft Fabric environments.

Acknowledgments

Generated by Claude Code
Built for Microsoft Fabric
Uses Power BI REST API
Leverages semantic-link library for Semantic Model integration

Quick Reference

Notebook Parameters

# Report configuration
workspace_id = ""
report_id = ""
output_format = "PDF"
static_params = "{}"
report_partitioning_column = "Customer"

# Source configuration
report_partitioning_source = "semantic_model"  # or "lakehouse", "json", "warehouse"

# Semantic Model
semantic_model_workspace_id = ""
semantic_model_dataset_id = ""
semantic_model_dax_query = "EVALUATE DISTINCT('Table'[Column])"

# Lakehouse
lakehouse_table = "parameter_config"
lakehouse_category = "CustomerList"
lakehouse_column = "ParameterValue"
lakehouse_filter = ""

# JSON
report_partitioning_values = "[]"

# Warehouse
warehouse_name = ""
warehouse_table = "dbo.ParameterConfig"
warehouse_column = "ParameterValue"
warehouse_category = ""

# Options
archive_to_onelake = "true"
max_retries = "3"
export_timeout_seconds = "600"
poll_interval_seconds = "5"

Pipeline Triggers

{
  "DailySchedule": "6:00 AM every day",
  "WeeklySchedule": "8:00 AM every Monday",
  "MonthlySchedule": "7:00 AM on 1st of month"
}

Ready to generate thousands of reports with ease! 🚀

This site is open source. Improve this page.

fabric-paginated-report-sch

Fabric Paginated Report Batch Executor

Features

Table of Contents

Quick Start

1. Upload the Notebook

2. Choose Your Parameter Source

3. Run a Test

Architecture

Execution Flow

Parameter Sources

1. Semantic Model (Recommended for Business Users)

2. Lakehouse (Recommended for Data Engineers)

3. JSON Array (Recommended for Testing)

4. Warehouse (Recommended for Enterprise)

Setup Instructions

Prerequisites

Step 1: Import Notebook

Step 2: Set Up Parameter Source

Option A: Semantic Model

Option B: Lakehouse

Option C: JSON

Option D: Warehouse

Step 3: Create Pipeline

Step 4: Test

Usage Examples

Example 1: Monthly Customer Reports (Semantic Model)

Example 2: Regional Reports (Lakehouse)

Example 3: Testing with JSON

Example 4: Enterprise Warehouse with RLS

Pipeline Configuration

Required Parameters

Source-Specific Parameters

Optional Parameters

Advanced Features

1. Multiple Parameter Loops

2. Conditional Parameter Lists

3. Dynamic File Naming

4. Error Handling and Retry

5. Monitoring and Notifications

6. Performance Tuning Parameters

Troubleshooting

Common Issues

1. No Parameter Values Loaded

2. Authentication Errors

3. Report Export Timeout

4. Slow Performance

Debug Mode

Best Practices

1. Parameter Source Selection

2. File Organization

3. Scheduling

4. Security

5. Maintenance

Support and Contributing

Getting Help

File Structure

Version History

License

Acknowledgments

Quick Reference

Notebook Parameters

Pipeline Triggers