langchain-ai/langgraph
1
0
Fork 0
mirror of https://github.com/langchain-ai/langgraph.git synced 2026-07-15 12:35:43 -04:00
Code Issues 314 Packages Projects Releases 17 Wiki Activity

Calling .astream inside of another graph's .astream_events will set stream_mode="value" regardless of config #304

New Issue
Closed
opened 2026-02-20 17:36:00 -05:00 by yindo · 4 comments
yindo commented 2026-02-20 17:36:00 -05:00
Owner
Copy Link
Copy Source

Originally created by @lucaslulucaslu on GitHub (Nov 6, 2024).

Checked other resources

  • I added a very descriptive title to this issue.
  • I searched the LangGraph/LangChain documentation with the integrated search.
  • I used the GitHub search to find a similar question and didn't find it.
  • I am sure that this is a bug in LangGraph/LangChain rather than my code.
  • I am sure this is better as an issue rather than a GitHub discussion, since this is a LangGraph bug and not a design question.

Example Code

cannot have code here due to security issue

Error Message and Stack Trace (if applicable)

No response

Description

I have a graph A streaming LLM tokens, I tested it and can confirm I get updates as streaming. Then I have another graph B which is invoked by B.astream_events, then invoke A.astream, I can only get values of every GraphState, no matter I set stream_mode to "value" or "update"

System Info

latest langgraph
mac
python 3.12

Originally created by @lucaslulucaslu on GitHub (Nov 6, 2024). ### Checked other resources - [X] I added a very descriptive title to this issue. - [X] I searched the [LangGraph](https://langchain-ai.github.io/langgraph/)/LangChain documentation with the integrated search. - [X] I used the GitHub search to find a similar question and didn't find it. - [X] I am sure that this is a bug in LangGraph/LangChain rather than my code. - [X] I am sure this is better as an issue [rather than a GitHub discussion](https://github.com/langchain-ai/langgraph/discussions/new/choose), since this is a LangGraph bug and not a design question. ### Example Code ```python cannot have code here due to security issue ``` ### Error Message and Stack Trace (if applicable) _No response_ ### Description I have a graph A streaming LLM tokens, I tested it and can confirm I get updates as streaming. Then I have another graph B which is invoked by B.astream_events, then invoke A.astream, I can only get values of every GraphState, no matter I set stream_mode to "value" or "update" ### System Info latest langgraph mac python 3.12
yindo closed this issue 2026-02-20 17:36:00 -05:00
yindo commented 2026-02-20 17:36:00 -05:00
Author
Owner
Copy Link
Copy Source

@vbarda commented on GitHub (Nov 6, 2024):

@lucaslulucaslu could you write some pseudocode to illustrate this issue better?

@vbarda commented on GitHub (Nov 6, 2024): @lucaslulucaslu could you write some pseudocode to illustrate this issue better?
yindo commented 2026-02-20 17:36:01 -05:00
Author
Owner
Copy Link
Copy Source

@lucaslulucaslu commented on GitHub (Nov 9, 2024):

Sure @vbarda

from langgraph.graph import StateGraph,START,END
from pydantic import BaseModel

class GraphState1(BaseModel):
    a: int = 0
    b: int = 0

def graph1_node1(state: GraphState1):
    a=state.a
    a+=1
    return {"a":a}
def graph1_node2(state: GraphState1):
    b=state.b
    b+=1
    return {"b":b}


graph1=StateGraph(GraphState1)
graph1.add_node(graph1_node1.__name__,graph1_node1)
graph1.add_node(graph1_node2.__name__,graph1_node2)
graph1.add_edge(START,graph1_node1.__name__)
graph1.add_edge(graph1_node1.__name__,graph1_node2.__name__)
graph1.add_edge(graph1_node2.__name__,END)
graph1_app= graph1.compile()

async for chunk in graph1_app.astream({"a":0,"b":0}):
    print(chunk)

This will output updates as expected:

{'graph1_node1': {'a': 1}}
{'graph1_node2': {'b': 1}}

However if we invoke this exactly same graph in another graph:

class GraphState2(BaseModel):
    c: int = 0
    d: int = 0

async def graph2_node1(state: GraphState2):
    async for chunk in graph1_app.astream({"a":0,"b":0},stream_mode="updates"):
        print(chunk)
    c=state.c
    c+=1
    return {"c":c}
def graph2_node2(state: GraphState2):
    d=state.d
    d+=1
    return {"d":d}

graph2=StateGraph(GraphState2)
graph2.add_node(graph2_node1.__name__,graph2_node1)
graph2.add_node(graph2_node2.__name__,graph2_node2)
graph2.add_edge(START,graph2_node1.__name__)
graph2.add_edge(graph2_node1.__name__,graph2_node2.__name__)
graph2.add_edge(graph2_node2.__name__,END)
graph2_app= graph2.compile()

async for chunk in graph2_app.astream_events({"c":0,"d":0},version="v2"):
    pass

graph2 does nothing but invoke graph1 in graph2_node1, and the print out result is:

{'a': 0, 'b': 0}
{'a': 1, 'b': 0}
{'a': 1, 'b': 1}

even I explicitly use async for chunk in graph1_app.astream({"a":0,"b":0},stream_mode="updates"): with stream_mode ="updates"

@lucaslulucaslu commented on GitHub (Nov 9, 2024): Sure @vbarda ``` from langgraph.graph import StateGraph,START,END from pydantic import BaseModel class GraphState1(BaseModel): a: int = 0 b: int = 0 def graph1_node1(state: GraphState1): a=state.a a+=1 return {"a":a} def graph1_node2(state: GraphState1): b=state.b b+=1 return {"b":b} graph1=StateGraph(GraphState1) graph1.add_node(graph1_node1.__name__,graph1_node1) graph1.add_node(graph1_node2.__name__,graph1_node2) graph1.add_edge(START,graph1_node1.__name__) graph1.add_edge(graph1_node1.__name__,graph1_node2.__name__) graph1.add_edge(graph1_node2.__name__,END) graph1_app= graph1.compile() async for chunk in graph1_app.astream({"a":0,"b":0}): print(chunk) ``` This will output updates as expected: ``` {'graph1_node1': {'a': 1}} {'graph1_node2': {'b': 1}} ``` However if we invoke this exactly same graph in another graph: ``` class GraphState2(BaseModel): c: int = 0 d: int = 0 async def graph2_node1(state: GraphState2): async for chunk in graph1_app.astream({"a":0,"b":0},stream_mode="updates"): print(chunk) c=state.c c+=1 return {"c":c} def graph2_node2(state: GraphState2): d=state.d d+=1 return {"d":d} graph2=StateGraph(GraphState2) graph2.add_node(graph2_node1.__name__,graph2_node1) graph2.add_node(graph2_node2.__name__,graph2_node2) graph2.add_edge(START,graph2_node1.__name__) graph2.add_edge(graph2_node1.__name__,graph2_node2.__name__) graph2.add_edge(graph2_node2.__name__,END) graph2_app= graph2.compile() async for chunk in graph2_app.astream_events({"c":0,"d":0},version="v2"): pass ``` graph2 does nothing but invoke graph1 in graph2_node1, and the print out result is: ``` {'a': 0, 'b': 0} {'a': 1, 'b': 0} {'a': 1, 'b': 1} ``` even I explicitly use `async for chunk in graph1_app.astream({"a":0,"b":0},stream_mode="updates"):` with stream_mode ="updates"
yindo commented 2026-02-20 17:36:01 -05:00
Author
Owner
Copy Link
Copy Source

@OneCoin123 commented on GitHub (May 11, 2025):

I met the same problem. Is there any method to stream a subgraph calling in update mode? Or it is simply restricted to use stream_mode='values' for now?

@OneCoin123 commented on GitHub (May 11, 2025): I met the same problem. Is there any method to stream a subgraph calling in update mode? Or it is simply restricted to use stream_mode='values' for now?
yindo commented 2026-02-20 17:36:01 -05:00
Author
Owner
Copy Link
Copy Source

@nfcampos commented on GitHub (May 23, 2025):

Fixed in #4806

@nfcampos commented on GitHub (May 23, 2025): Fixed in #4806
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
agents
awaiting author revision
bug
change
chat models
checkpointer
ci
code quality
code quality
dependencies
documentation
drawing
duplicate
enhancement
good first issue
help wanted
invalid
maintainer
meta
missing information
not maintained by langchain
open-swe-auto
open-swe-max
open-swe-max-auto
pending
performance
pull-request
Mirrored from GitHub Pull Request
 question
requires design
sdk
server
stale
studio
tests
type checking
usage question
v0.5
v1.1
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langchain-ai/langgraph#304
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?