Failure Workflow

A processing pipeline fails mid-execution. The system must automatically clean up resources and notify the team without any manual trigger -- the failure handler must receive the original workflow's ID, type, and failure reason.

Workflow

fw_process ──(fails when shouldFail=true)──> failureWorkflow: error_handler_wf
                                                     │
                                              fw_cleanup ──> fw_notify_failure

The primary workflow main_with_failure_handler runs fw_process with retryCount = 0 and failureWorkflow = "error_handler_wf". The error handler workflow error_handler_wf chains fw_cleanup then fw_notify_failure, both configured with retryCount = 2 and responseTimeoutSeconds = 30.

Workers

ProcessWorker (fw_process) -- reads shouldFail from task input using a toBoolean() helper that handles Boolean, String, and null types. When true, returns FAILED with reasonForIncompletion = "Intentional failure: shouldFail was true" and outputs status = "FAILED". Otherwise returns status = "SUCCESS".

CleanupWorker (fw_cleanup) -- receives failedWorkflowId, failedWorkflowType, and reason from the error handler workflow's input parameters. Returns cleaned = true and message = "Cleanup completed successfully".

NotifyFailureWorker (fw_notify_failure) -- receives failedWorkflowId, cleaned status from the cleanup step, and the original reason. Returns notified = true and message = "Failure notification sent".

Workflow Output

The workflow produces status, message as output parameters, capturing the result of each pipeline stage for downstream consumers and observability.

Task Configuration

• fw_process: retryCount=0, retryLogic=FIXED, retryDelaySeconds=?, timeoutSeconds=60, responseTimeoutSeconds=30
• fw_cleanup: retryCount=2, retryLogic=FIXED, retryDelaySeconds=?, timeoutSeconds=60, responseTimeoutSeconds=30
• fw_notify_failure: retryCount=2, retryLogic=FIXED, retryDelaySeconds=?, timeoutSeconds=60, responseTimeoutSeconds=30

These settings are declared in task-defs.json and apply independently to each task, controlling retry behavior, timeout detection, and backoff strategy without any changes to worker code.

Project Structure

This example contains 3 worker implementations in src/main/java/*/workers/, the workflow definition in src/main/resources/workflow.json, and integration tests in src/test/. The workflow main_with_failure_handler defines 1 task with input parameters shouldFail and a timeout of 60 seconds.

Tests

4 tests verify successful processing, failure triggering, cleanup execution, and notification delivery with correct failure context propagation.

Running

See RUNNING.md for setup and execution instructions.