Testing Workflows

Thoroughly test your workflows before deployment to ensure they work correctly and handle edge cases. The workflow builder provides powerful testing and debugging tools.

Test Run Overview

Test runs allow you to execute workflows in a controlled environment without affecting production data or triggering external systems.

Starting a Test Run

1. Click Test Run button in the workflow builder toolbar
2. Provide test input data (if required by trigger)
3. Watch real-time execution on the canvas
4. Review results and logs

Test Run Features

Feature	Description
Real-Time Visualization	Nodes light up during execution
Node Output Inspection	View data at each step
Execution Logs	Console output and errors
Performance Metrics	Duration per node
Debug Mode	Verbose logging

Providing Test Input

Different trigger types require different test inputs.

Webhook Trigger

Simulate an HTTP request:

{
  "method": "POST",
  "headers": {
    "Content-Type": "application/json",
    "X-Custom-Header": "value"
  },
  "body": {
    "event": "user.created",
    "user_id": "12345",
    "email": "test@example.com"
  },
  "query": {
    "source": "api"
  }
}

Schedule Trigger

No input required - uses current timestamp:

{
  "scheduledTime": "2024-01-15T09:00:00Z",
  "actualTime": "2024-01-15T09:00:01Z"
}

REST API Trigger

Simulate API request:

{
  "method": "POST",
  "path": "/api/process",
  "headers": {
    "Authorization": "Bearer test-token"
  },
  "body": {
    "input": "data to process"
  }
}

Form Trigger

Simulate form submission:

{
  "formData": {
    "name": "John Smith",
    "email": "john@example.com",
    "message": "This is a test submission"
  },
  "metadata": {
    "submittedAt": "2024-01-15T10:30:00Z",
    "userAgent": "Mozilla/5.0...",
    "ipAddress": "192.168.1.1"
  }
}

Real-Time Execution Visualization

During test runs, the canvas provides visual feedback:

Node States

State	Appearance	Meaning
Pending	Gray	Not yet executed
Running	Blue/Pulsing	Currently executing
Success	Green	Completed successfully
Error	Red	Failed with error
Skipped	Yellow	Skipped by condition

Execution Flow

• Connection Highlights: Active data flow shown in blue
• Progress Indicator: Shows current execution position
• Parallel Execution: Multiple nodes light up simultaneously
• Completion Status: Final state for each node

Inspecting Node Output

Click any node to view its execution details:

Output Data Tab

View the data produced by the node:

{
  "status": "success",
  "data": {
    "id": "12345",
    "processed": true,
    "timestamp": "2024-01-15T10:30:00Z"
  },
  "metadata": {
    "duration": 245,
    "nodeId": "node_abc123"
  }
}

Features:

• JSON formatting with syntax highlighting
• Expand/collapse nested objects
• Copy data to clipboard
• Search within output

Execution Logs Tab

View console output and logs:

[10:30:00.123] Starting node execution
[10:30:00.156] Fetching data from API
[10:30:00.401] API response received (200 OK)
[10:30:00.405] Parsing response data
[10:30:00.410] Node execution complete

Log Levels:

• INFO: General information
• DEBUG: Detailed execution steps
• WARN: Warnings and issues
• ERROR: Error messages and stack traces

Performance Tab

View timing and resource metrics:

Metric	Value
Start Time	10:30:00.123
End Time	10:30:00.410
Duration	287ms
CPU Usage	15%
Memory Usage	45MB

Debug Mode

Enable debug mode for verbose logging and detailed execution information.

Enabling Debug Mode

1. Click Settings in toolbar
2. Toggle Debug Mode on
3. Run test again
4. Review enhanced logs

Debug Information

Enhanced Logs:

[DEBUG] Node: apiCall
[DEBUG] Input variables:
{
  "userId": "12345",
  "endpoint": "/users/12345"
}
[DEBUG] Executing HTTP request
[DEBUG] Request headers: {...}
[DEBUG] Request body: {...}
[DEBUG] Response status: 200
[DEBUG] Response headers: {...}
[DEBUG] Response body: {...}
[DEBUG] Parsing response...
[DEBUG] Output mapped to: {...}

Variable Inspection:

• View all variables at each node
• See variable transformations
• Track data flow through pipeline
• Inspect intermediate values

Testing Best Practices

Test Data Preparation

1. Realistic Data: Use production-like test data
2. Edge Cases: Test boundary conditions
3. Invalid Data: Test error handling
4. Various Formats: Test different input formats

Example Test Cases:

// Happy path
{
  "userId": "12345",
  "email": "valid@example.com",
  "age": 30
}

// Edge case - minimum values
{
  "userId": "1",
  "email": "a@b.c",
  "age": 18
}

// Edge case - maximum values
{
  "userId": "999999999",
  "email": "very.long.email.address@subdomain.example.com",
  "age": 120
}

// Invalid data
{
  "userId": null,
  "email": "invalid-email",
  "age": -5
}

// Missing fields
{
  "userId": "12345"
  // email and age missing
}

Testing Scenarios

1. Success Path

• All nodes execute successfully
• Data flows as expected
• Output format is correct

2. Error Handling

• API failures
• Invalid data formats
• Network timeouts
• Authentication errors

3. Conditional Logic

• Test all branches
• Verify routing conditions
• Check default paths

4. Edge Cases

• Empty arrays
• Null values
• Very large datasets
• Special characters

Iterative Testing

1. Initial Test: Run with basic data
2. Review Results: Check output and logs
3. Fix Issues: Update node configurations
4. Retest: Run again with fixes
5. Expand Tests: Add edge cases
6. Final Validation: Test all scenarios

Common Testing Issues

Issue: Node Not Executing

Symptoms:

• Node remains gray during test
• No output data
• No logs

Causes:

• Missing connection from previous node
• Conditional logic skipping node
• Disabled node

Solutions:

• Verify node connections
• Check conditional expressions
• Ensure node is enabled

Issue: Incorrect Output

Symptoms:

• Output doesn't match expected format
• Missing or incorrect data
• Data transformation errors

Causes:

• Wrong variable references
• Incorrect data mapping
• Type conversion issues

Solutions:

• Review variable syntax: {{ nodeName.field }}
• Check data types (string vs. number)
• Validate mapping expressions
• Use debug mode to inspect variables

Issue: API Call Failures

Symptoms:

• Node shows error state
• HTTP error codes
• Timeout errors

Causes:

• Invalid credentials
• Wrong endpoint URL
• Network issues
• Rate limiting

Solutions:

• Verify data source credentials
• Check endpoint URLs
• Add retry logic
• Test with API testing tool first

Issue: Slow Execution

Symptoms:

• Long test run duration
• Individual nodes taking too long
• Timeout errors

Causes:

• Large data processing
• Sequential dependencies
• Inefficient queries
• External API latency

Solutions:

• Optimize data transformations
• Use parallel execution
• Add caching
• Batch operations
• Increase timeout values

Execution Timeline

View the execution timeline to analyze performance:

Waterfall Chart

Visualizes node execution order and duration:

Trigger          |▓|
  └─ API Call    |  ▓▓▓▓▓|
      └─ Parse   |       ▓|
          └─ AI  |        ▓▓▓▓▓▓▓▓▓▓|
              └─ DB|                ▓▓|
└─ Parallel Node |  ▓▓▓|

Features:

• Identify bottlenecks
• See parallel execution
• Measure node duration
• Detect slow operations

Performance Analysis

Node	Start	Duration	% of Total
Trigger	0ms	2ms	0.1%
API Call	2ms	450ms	22.5%
Parse	452ms	5ms	0.25%
AI Gateway	457ms	1500ms	75%
MongoDB	1957ms	45ms	2.25%

Optimization Tips:

• Focus on nodes with highest % of total time
• Look for unnecessary sequential dependencies
• Consider caching for repeated operations
• Batch operations where possible

Testing Checklist

Before deploying to production, verify:

Functionality

• All nodes execute successfully
• Data flows correctly through pipeline
• Output format matches requirements
• Error handling works as expected

Data Validation

• Input validation is working
• Output data is correct
• Data transformations are accurate
• Edge cases are handled

Error Handling

• Network failures are handled
• Invalid data is caught
• Retry logic works
• Error messages are clear

Performance

• Execution time is acceptable
• No unnecessary sequential operations
• Parallel execution is used where possible
• Timeouts are appropriate

Security

• Credentials are not exposed
• Input is sanitized
• Output doesn't leak sensitive data
• Authentication is working

Advanced Testing

Load Testing

Test workflow performance under load:

1. Create Test Script: Generate multiple requests
2. Gradual Increase: Start with low volume
3. Monitor Metrics: Watch performance degradation
4. Identify Limits: Find breaking points

Integration Testing

Test with real external systems:

1. Test Environment: Use sandbox/test accounts
2. Real Data: Use production-like data
3. End-to-End: Test complete workflow
4. Verify Results: Check external system state

Regression Testing

Test after making changes:

1. Baseline: Record initial test results
2. Make Changes: Update workflow
3. Retest: Run same tests
4. Compare: Verify no regressions

Next Steps

Once your workflow is thoroughly tested:

• Deploy to production
• Monitor workflow executions
• Set up alerts
• Review workflow examples

Test Thoroughly

Invest time in testing before deployment. Well-tested workflows are more reliable and easier to maintain.