Skip to content

LLM-Driven Extraction of Unstructured Data — Built for API Deployments & ETL Pipeline Workflows

UN-2640 [FIX] Resolve empty task_id causing 404 status API errors (#1453)

ali8000a8003bd8
diff --git a/backend/workflow_manager/workflow_v2/exceptions.py b/backend/workflow_manager/workflow_v2/exceptions.py
index 6f52081f8..d6b10ede3 100644
--- a/backend/workflow_manager/workflow_v2/exceptions.py
+++ b/backend/workflow_manager/workflow_v2/exceptions.py
@@ -26,11 +26,6 @@ class ExecutionDoesNotExistError(APIException):
default_detail = "Execution does not exist."
-class TaskDoesNotExistError(APIException):
- status_code = 404
- default_detail = "Task does not exist"
-
-
class DuplicateActionError(APIException):
status_code = 400
default_detail = "Action is running"
diff --git a/backend/workflow_manager/workflow_v2/execution.py b/backend/workflow_manager/workflow_v2/execution.py
index 34dd2931e..3de4cb6c1 100644
--- a/backend/workflow_manager/workflow_v2/execution.py
+++ b/backend/workflow_manager/workflow_v2/execution.py
@@ -404,15 +404,22 @@ class WorkflowExecutionServiceHelper(WorkflowExecutionService):
@staticmethod
def update_execution_task(execution_id: str, task_id: str) -> None:
try:
+ logger.info(
+ f"Updating task_id for execution {execution_id} with task_id {task_id}"
+ )
if not task_id:
logger.warning(
- f"task_id: '{task_id}' is NULL / empty for "
- f"execution_id: {execution_id}, expected to have a UUID"
+ f"Skipped setting task_id for execution {execution_id} since its NULL/empty"
)
+ return # Don't save empty task_id
+
execution = WorkflowExecution.objects.get(pk=execution_id)
# TODO: Review if status should be updated to EXECUTING
execution.task_id = task_id
- execution.save()
+ execution.save(update_fields=["task_id"])
+ logger.info(
+ f"Successfully set task_id '{task_id}' for execution {execution_id}"
+ )
except WorkflowExecution.DoesNotExist:
logger.error(f"execution doesn't exist {execution_id}")
diff --git a/backend/workflow_manager/workflow_v2/workflow_helper.py b/backend/workflow_manager/workflow_v2/workflow_helper.py
index b0b7ac2b6..dc7409432 100644
--- a/backend/workflow_manager/workflow_v2/workflow_helper.py
+++ b/backend/workflow_manager/workflow_v2/workflow_helper.py
@@ -55,7 +55,6 @@ from workflow_manager.workflow_v2.enums import (
)
from workflow_manager.workflow_v2.exceptions import (
InvalidRequest,
- TaskDoesNotExistError,
WorkflowDoesNotExistError,
WorkflowExecutionError,
WorkflowExecutionNotExist,
@@ -343,16 +342,12 @@ class WorkflowHelper:
execution_id (str): workflow execution id
Raises:
- TaskDoesNotExistError: Not found exception
ExecutionDoesNotExistError: If execution is not found
Returns:
ExecutionResponse: _description_
"""
execution: WorkflowExecution = WorkflowExecution.objects.get(id=execution_id)
- if not execution.task_id:
- raise TaskDoesNotExistError(f"No task ID found for execution: {execution_id}")
-
task_result = None
result_acknowledged = execution.result_acknowledged
# Prepare the initial response with the task's current status and result.
@@ -473,15 +468,32 @@ class WorkflowHelper:
},
queue=queue,
)
+
+ # Log task_id for debugging
logger.info(
- f"[{org_schema}] Job '{async_execution}' has been enqueued for "
- f"execution_id '{execution_id}', '{len(hash_values_of_files)}' files"
+ f"[{org_schema}] AsyncResult created with task_id: '{async_execution.id}' "
+ f"(type: {type(async_execution.id).__name__})"
)
+
workflow_execution: WorkflowExecution = WorkflowExecution.objects.get(
id=execution_id
)
- workflow_execution.task_id = async_execution.id
- workflow_execution.save()
+
+ # Handle empty task_id gracefully using existing validation logic
+ if not async_execution.id:
+ logger.warning(
+ f"[{org_schema}] Celery returned empty task_id for execution_id '{execution_id}'. "
+ )
+ # Continue without setting task_id - execution can still complete
+ else:
+ # Use existing method to handle task_id setting with validation
+ WorkflowExecutionServiceHelper.update_execution_task(
+ execution_id=execution_id, task_id=async_execution.id
+ )
+ logger.info(
+ f"[{org_schema}] Job '{async_execution.id}' has been enqueued for "
+ f"execution_id '{execution_id}', '{len(hash_values_of_files)}' files"
+ )
execution_status = workflow_execution.status
if timeout > -1:
while not ExecutionStatus.is_completed(execution_status) and timeout > 0: