Skip to main content
Performance test apps using the Business Central Performance Toolkit (BCPT) allow you to measure execution times, SQL statements, and other performance metrics. AL-Go automates BCPT test execution and provides detailed performance reports with baseline comparisons.

​
Prerequisites

Before adding a performance test app:
  • Complete the basic AL-Go for GitHub setup
  • Understand BCPT framework basics
  • Have a working CI/CD pipeline

​
Create a Performance Test App

1

Run the Create Performance Test App workflow

Navigate to Actions in your repository, select Create a new performance test app, then click Run workflow.Provide the required information:
  • Name: Name of your performance test app
  • Publisher: Your publisher name
  • ID Range: Object ID range for the test app
Click Run workflow to create the performance test app.
If workflows are not allowed to create pull requests due to GitHub organization settings, you can create the PR manually by following the link in the workflow annotation.
2

Review and merge the pull request

Once the workflow completes, navigate to Pull Requests.Inspect the changes, which include:
  • Performance test app structure
  • Sample BCPT test codeunits
  • BCPT suite configuration
Merge the pull request when ready.
3

Monitor the CI/CD workflow

After merging, a CI/CD workflow automatically starts.Navigate to Actions to watch the workflow execute. This workflow will:
  • Build your apps and performance test app
  • Run the BCPT tests
  • Generate performance test results
4

Review BCPT test results

When the workflow completes successfully, check the build artifacts.You’ll find BCPT Test Results as one of the artifacts. Download and open it to view:
  • Execution duration for each test
  • Number of SQL statements
  • SQL read/write operations
  • Performance metrics in JSON and table format

​
Understanding BCPT Results

The BCPT Test Results artifact contains detailed performance data: 
{
  "TestName": "Customer Lookup",
  "Duration": 125,
  "NumberOfSQLStatements": 8,
  "NumberOfSQLReads": 15,
  "NumberOfSQLWrites": 0
}

Duration

Execution time in milliseconds for the test operation

SQL Statements

Total number of SQL statements executed during the test

SQL Reads

Number of database read operations performed

SQL Writes

Number of database write operations performed

​
Set Up Performance Baselines

To compare performance across builds, establish a baseline by uploading a bcptBaseLine.json file.
1

Get initial results

After your first successful BCPT run, download the BCPT Test Results artifact.The results viewer will indicate that you need to add a bcptBaseLine.json file to enable baseline comparison.
2

Create baseline file

Use the results from your initial run to create bcptBaseLine.json.Upload this file to your project root (the repository root for single-project repositories).
{
  "TestName": "Customer Lookup",
  "Duration": 125,
  "NumberOfSQLStatements": 8,
  "NumberOfSQLReads": 15,
  "NumberOfSQLWrites": 0
}
3

Run CI/CD with baseline

Commit and push the bcptBaseLine.json file.The next CI/CD workflow will compare new results against the baseline, showing:
  • Diff values for each metric
  • Negative numbers indicate improvements (faster execution, fewer SQL operations)
  • Positive numbers indicate performance regression

​
Configure Performance Thresholds

Set thresholds to automatically fail builds when performance degrades beyond acceptable levels.
1

Create thresholds file

Create a bcptThresholds.json file in the same location as your baseline file (project root).
{
  "durationWarning": 0,
  "durationError": 1
}
You can also specify thresholds in project settings using the bcptThresholds setting. See Settings Documentation for details.
2

Understand threshold values

Threshold values represent percentage changes from baseline:
  • durationWarning: 0 - Warns if duration increases at all
  • durationError: 1 - Fails if duration increases by more than 1%
  • Similar thresholds can be set for SQL statements, reads, and writes
3

Test threshold enforcement

Commit and push the bcptThresholds.json file.The CI/CD workflow will now:
  • Show warnings when performance crosses warning thresholds
  • Fail the build when error thresholds are exceeded
  • Display threshold violations in the workflow summary

​
BCPT Threshold Options

You can configure multiple threshold types: 
{
  "durationWarning": 5,
  "durationError": 10,
  "numberOfSQLStatementsWarning": 5,
  "numberOfSQLStatementsError": 10,
  "numberOfSQLReadsWarning": 5,
  "numberOfSQLReadsError": 10,
  "numberOfSQLWritesWarning": 5,
  "numberOfSQLWritesError": 10
}

​
Best Practices

​
Choose Representative Tests

Focus on critical scenarios: Performance test the most important and frequently used features in your application, not every possible operation.
  • Test common user workflows (e.g., creating sales orders, posting documents)
  • Include scenarios that access large datasets
  • Test operations known to be performance-sensitive
  • Consider testing during peak load conditions

​
Maintain Stable Baselines

  • Update baselines when you intentionally improve performance
  • Don’t update baselines to hide regressions
  • Document why baselines changed in commit messages
  • Keep historical baselines for comparison

​
Set Realistic Thresholds

{
  "durationWarning": 5,     // Warn at 5% regression
  "durationError": 15       // Fail at 15% regression
}
  • Start with lenient thresholds and tighten over time
  • Account for normal variation in test execution
  • Use warnings to catch trends before they become problems
  • Set error thresholds to prevent significant regressions

Track Over Time

Download and archive BCPT results from each build to identify performance trends across releases.

Compare Branches

Run BCPT tests on feature branches to catch performance issues before merging.

Document Changes

When performance characteristics change, document the reasons in release notes.

Investigate Anomalies

Sudden performance changes deserve investigation even if they don’t breach thresholds.

​
Example BCPT Test

codeunit 50100 "BCPT Customer Operations"
{
    // Simulates customer lookup and read operations
    procedure RunCustomerLookup()
    var
        Customer: Record Customer;
    begin
        if Customer.FindSet() then
            repeat
                // Perform typical operations
            until (Customer.Next() = 0);
    end;
}

​
Next Steps

Add Test Apps

Create functional test apps to validate business logic

Development Environments

Set up online development environments for rapid testing

Build docs developers (and LLMs) love

Get started for freeTalk to us