Workflow trace error

[wy96f]

Self Checks

• This is only for bug report, if you would like to ask a question, please head to Discussions.
• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
• Please do not modify this template :) and fill in all the required fields.

Dify version

0.14.2

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

1. configure langfuse
2. execute workflow

✔️ Expected Behavior

trace successfully sended to langfuse

❌ Actual Behavior

2024-12-26 [2024-12-26 11:54:50,469][ERROR][trace_manager_timer_2024-12-26 03:54:45][ops_trace_manager.py:739] Error processing trace tasks

2024-12-26 Traceback (most recent call last):

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 737, in run

2024-12-26 self.send_to_celery(tasks)

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 753, in send_to_celery

2024-12-26 trace_info = task.execute()

2024-12-26 ^^^^^^^^^^^^^^

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 330, in execute

2024-12-26 return self.preprocess()

2024-12-26 ^^^^^^^^^^^^^^^^^

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 352, in preprocess

2024-12-26 return preprocess_map.get(self.trace_type, lambda: None)()

2024-12-26 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 335, in

2024-12-26 TraceTaskName.WORKFLOW_TRACE: lambda: self.workflow_trace(

2024-12-26 ^^^^^^^^^^^^^^^^^^^^

2024-12-26 File "/app/api/core/ops/ops_trace_manager.py", line 363, in workflow_trace

2024-12-26 db.session.refresh(workflow_run)

2024-12-26 File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/scoping.py", line 1767, in refresh

2024-12-26 return self._proxied.refresh(

2024-12-26 ^^^^^^^^^^^^^^^^^^^^^^

2024-12-26 File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 3143, in refresh

2024-12-26 self._expire_state(state, attribute_names)

2024-12-26 File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 3269, in _expire_state

2024-12-26 self._validate_persistent(state)

2024-12-26 File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 4138, in _validate_persistent

2024-12-26 raise sa_exc.InvalidRequestError(

2024-12-26 sqlalchemy.exc.InvalidRequestError:Instance '<WorkflowRun at 0x7fb237659bb0>' is not persistent within this Session

2024-12-26 [2024-12-26 11:55:08,350][DEBUG][Dummy-40][_trace.py:47] close.started

2024-12-26 [2024-12-26 11:55:08,350][DEBUG][Dummy-40][_trace.py:47] close.complete

2024-12-26 [2024-12-26 11:56:45,503][DEBUG][Dummy-91][_trace.py:47] close.started

2024-12-26 [2024-12-26 11:56:45,503][DEBUG][Dummy-91][_trace.py:47] close.complete

[dosubot]

I found a similar open issue related to a workflow trace error in Dify. The error message "Instance '<Workflow at 0x7f7800591750>' is not persistent within this Session" suggests that the Workflow instance is not associated with the current SQLAlchemy session when db.session.refresh() is called. Here are some steps that were suggested to resolve this issue:

1. Ensure the Instance is Added to the Session: Before calling db.session.refresh(), make sure that the Workflow instance is added to the session using db.session.add(workflow_instance).

2. Check for Session Expiry: If the session was closed or expired before the refresh operation, you might need to reopen or reattach the instance to a new session.

3. Verify the Instance State: Use db.session.is_modified(workflow_instance) to check if the instance is in a state that can be refreshed. If it returns False, the instance might not be in a persistent state.

4. Review Transaction Boundaries: Ensure that the session is not being prematurely committed or rolled back, which could detach the instance from the session.

By ensuring that the Workflow instance is properly managed within the session lifecycle, you should be able to resolve this error ^[1].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Other}

[laipz8200]

fixed in #12071

[mgmorcos]

Still seeing this issue on main:

sqlalchemy.exc.InvalidRequestError: Instance '<WorkflowRun at 0x7fba188049e0>' is not persistent within this Session

[laipz8200]

Still seeing this issue on main:

sqlalchemy.exc.InvalidRequestError: Instance '<WorkflowRun at 0x7fba188049e0>' is not persistent within this Session

Hi, could you please provide more detailed information, such as the trace log? We want to know where this error happened.

[mgmorcos]

Hi @laipz8200 👋 ,

Still seeing on 0.15.0. This is from the log of the api container:

api-1  | 2025-01-12 06:22:14,119.119 ERROR [trace_manager_timer_2025-01-12 06:22:09] [ops_trace_manager.py:739] - Error processing trace tasks
api-1  | Traceback (most recent call last):
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 737, in run
api-1  |     self.send_to_celery(tasks)
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 753, in send_to_celery
api-1  |     trace_info = task.execute()
api-1  |                  ^^^^^^^^^^^^^^
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 330, in execute
api-1  |     return self.preprocess()
api-1  |            ^^^^^^^^^^^^^^^^^
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 352, in preprocess
api-1  |     return preprocess_map.get(self.trace_type, lambda: None)()
api-1  |            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 335, in <lambda>
api-1  |     TraceTaskName.WORKFLOW_TRACE: lambda: self.workflow_trace(
api-1  |                                           ^^^^^^^^^^^^^^^^^^^^
api-1  |   File "/app/api/core/ops/ops_trace_manager.py", line 363, in workflow_trace
api-1  |     db.session.refresh(workflow_run)
api-1  |   File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/scoping.py", line 1767, in refresh
api-1  |     return self._proxied.refresh(
api-1  |            ^^^^^^^^^^^^^^^^^^^^^^
api-1  |   File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 3143, in refresh
api-1  |     self._expire_state(state, attribute_names)
api-1  |   File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 3269, in _expire_state
api-1  |     self._validate_persistent(state)
api-1  |   File "/app/api/.venv/lib/python3.12/site-packages/sqlalchemy/orm/session.py", line 4138, in _validate_persistent
api-1  |     raise sa_exc.InvalidRequestError(
api-1  | sqlalchemy.exc.InvalidRequestError: Instance '<WorkflowRun at 0x7fe6f656a000>' is not persistent within this Session

TY! -M

[laipz8200]

Hi, @mgmorcos. The logs you prividered seems not belongs to the v0.15.0 version, please check you docker image.