Skip to main content

Advanced Features

The BankLingo Process Engine includes advanced BPMN 2.0 capabilities implemented across Phases 1-5.

Overview​

These advanced features extend the core BPMN functionality with production-ready implementations:

PhaseFeatureStatusDescription
Phase 1Multi-Instance Subprocess✅ ProductionExecute subprocess multiple times in parallel or sequence
Phase 2Async Boundaries✅ ProductionReturn API response immediately, continue in background
Phase 3Callbacks✅ ProductionWait for subprocess or external system completion
Phase 4Timer Events✅ ProductionSchedule tasks, timeouts, and repeating timers
Phase 5Error Handling✅ ProductionStructured error management with recovery patterns

Phase 1: Multi-Instance Execution​

Process collections of data in parallel or sequence:

Phase 2: Async Boundaries​

Enable background processing with immediate API responses:

Phase 3: Callbacks​

Wait for external events and subprocess completion:

Phase 4: Timer Events​

Schedule and delay process execution:

Phase 5: Error Handling​

Structured error management with recovery:

Feature Comparison​

When to Use Each Feature​

ScenarioFeatureWhy
Process 100 ordersMulti-InstanceParallel execution
Generate reportAsync BoundaryImmediate response
Wait for payment gatewayCallbacksExternal system integration
Enforce 48h SLATimer EventsTimeout boundary event
Handle API failuresError HandlingStructured retry logic

Learning Path​

Beginner Level​

  1. Start with Timer Events - Easiest to understand
  2. Try Async Boundaries - Simple configuration
  3. Learn Error Handling - Critical for production

Intermediate Level​

  1. Master Multi-Instance - Batch processing
  2. Implement Callbacks - External integration
  3. Combine features for complex workflows

Advanced Level​

  1. Build fault-tolerant workflows with error recovery
  2. Optimize performance with async boundaries
  3. Create orchestration patterns with callbacks

Integration Examples​

Example 1: Batch Processing with Error Handling​

<!-- Multi-instance subprocess with error boundaries -->
<bpmn:subProcess id="ProcessOrders" name="Process Each Order">
  <bpmn:multiInstanceLoopCharacteristics isSequential="false">
    <custom:properties>
      <custom:property name="collection" value="context.orders"/>
      <custom:property name="elementVariable" value="currentOrder"/>
    </custom:properties>
  </bpmn:multiInstanceLoopCharacteristics>
  
  <bpmn:scriptTask id="ProcessOrder" name="Process Order">
    <!-- Error handling integrated -->
  </bpmn:scriptTask>
  
  <!-- Boundary error event for individual order failures -->
  <bpmn:boundaryEvent id="OrderError" attachedToRef="ProcessOrder">
    <bpmn:errorEventDefinition/>
  </bpmn:boundaryEvent>
</bpmn:subProcess>

Example 2: SLA Enforcement with Escalation​

<!-- User task with timer boundary -->
<bpmn:userTask id="ApproveRequest" name="Manager Approval">
  <!-- ... -->
</bpmn:userTask>

<!-- 48-hour timeout -->
<bpmn:boundaryEvent id="Timeout" attachedToRef="ApproveRequest" cancelActivity="true">
  <bpmn:timerEventDefinition>
    <bpmn:timeDuration>P2D</bpmn:timeDuration>
  </bpmn:timerEventDefinition>
</bpmn:boundaryEvent>

<!-- Escalation on timeout -->
<bpmn:scriptTask id="Escalate" name="Escalate to Director">
  <!-- ... -->
</bpmn:scriptTask>

Example 3: Background Report with Callback​

<!-- Async boundary for immediate response -->
<bpmn:scriptTask id="GenerateReport" name="Generate Report" camunda:asyncBefore="true">
  <bpmn:extensionElements>
    <custom:properties>
      <custom:property name="asyncBoundary" value="true"/>
      <custom:property name="asyncBoundaryResponse"><![CDATA[{
        "jobId": "context.reportJobId",
        "status": "processing"
      }]]></custom:property>
    </custom:properties>
  </bpmn:extensionElements>
  <!-- Long-running operation -->
</bpmn:scriptTask>

<!-- Callback when external service completes -->
<bpmn:receiveTask id="WaitForCompletion" name="Wait for Upload">
  <!-- Callback mechanism -->
</bpmn:receiveTask>

Architecture Integration​

All advanced features integrate seamlessly with the core engine:

Process Execution
    ↓
Task Detection
    ↓
┌───────────────────────────────┐
│ Feature Detection             │
│ - Multi-instance?             │
│ - Async boundary?             │
│ - Timer event?                │
│ - Error boundary?             │
└───────────┬───────────────────┘
            ↓
Feature-Specific Execution
    ↓
State Management
    ↓
Continue or Pause

Performance Characteristics​

FeatureOverheadScalabilityBest For
Multi-InstanceLowLinear< 1000 items
Async BoundaryMinimalHighAny long task
CallbacksMinimalHighExternal waits
Timer EventsLowHighAny delays
Error HandlingMinimalHighAll tasks

Database Impact​

Phase 3: CallbackRegistry Table​

CREATE TABLE CallbackRegistry (
    Id BIGINT PRIMARY KEY,
    ProcessInstanceId NVARCHAR(200),
    CallbackType NVARCHAR(50),
    Status NVARCHAR(50)
);

Phase 4: ProcessTimers Table​

CREATE TABLE ProcessTimers (
    Id BIGINT PRIMARY KEY,
    ProcessInstanceId NVARCHAR(200),
    TimerEventId NVARCHAR(200),
    NextFireTime DATETIME,
    Status NVARCHAR(50)
);

Phase 5: Error Context​

Stored in ProcessState JSON:

{
  "errorHistory": [
    {
      "errorCode": "API_ERROR",
      "errorMessage": "Service unavailable",
      "timestamp": "2026-01-10T10:30:00Z"
    }
  ]
}

Next Steps​

  1. Choose a feature based on your use case
  2. Read the detailed documentation for that feature
  3. Try the examples in the documentation
  4. Combine features for complex workflows

Version: 2.0 (Phases 1-5)
Last Updated: January 2026