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

If the state of the parent graph is updated, then the state of the subgraph will be lost #633

New Issue
Closed
opened 2026-02-20 17:41:02 -05:00 by yindo · 5 comments
yindo commented 2026-02-20 17:41:02 -05:00
Owner
Copy Link
Copy Source

Originally created by @egortaran on GitHub (May 19, 2025).

Checked other resources

  • This is a bug, not a usage question. For questions, please use GitHub Discussions.
  • I added a clear and detailed title that summarizes the issue.
  • I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example).
  • I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue.

Example Code

from typing import TypedDict

from langchain_core.runnables import RunnableConfig
from langgraph.checkpoint.memory import MemorySaver
from langgraph.constants import START
from langgraph.graph import END, StateGraph


class State(TypedDict):
    foo: bool
    bar: bool


# Preparing subgraph
def subgraph_node_1(state: State) -> State:
    state["foo"] = True
    print("Executing `subgraph_node_1`")
    return state


def subgraph_node_2(state: State) -> State:
    print("Executing `subgraph_node_2`")
    return state


subgraph_builder = StateGraph(State)
subgraph_builder.add_node("subgraph_node_1", subgraph_node_1)
subgraph_builder.add_node("subgraph_node_2", subgraph_node_2)

subgraph_builder.add_edge(START, "subgraph_node_1")
subgraph_builder.add_edge("subgraph_node_1", "subgraph_node_2")
subgraph_builder.add_edge("subgraph_node_2", END)

subgraph = subgraph_builder.compile(
    interrupt_after=[
        "subgraph_node_1",
        "subgraph_node_2",
    ],
)

# Preparing parent graph
builder = StateGraph(State)


def node1(state: State) -> State:
    print("Executing `node1`")
    return state


def node3(state: State) -> State:
    print("Executing `node3`")
    return state


builder.add_node("node1", node1)
builder.add_node("node2", subgraph)
builder.add_node("node3", node3)

builder.add_edge(START, "node1")
builder.add_edge("node1", "node2")
builder.add_edge("node2", "node3")
builder.add_edge("node3", END)


# Graph executing
checkpointer = MemorySaver()
config = RunnableConfig(configurable={"thread_id": "1"})
graph = builder.compile(
    checkpointer=checkpointer,
    interrupt_after=["node1", "node3",],
)

for event in graph.stream({"foo": False, "bar": False}, config=config, stream_mode="updates", subgraphs=True):
    print(event)

print("*" * 10, "\n\n")


for event in graph.stream(None, config, subgraphs=True):
    print(event)

print("*" * 10, "\n\n")

# Update graph state
new_config = graph.update_state(config, {"bar": True})

for event in graph.stream(None, new_config, subgraphs=True):
    print(event)

print("*" * 10, "\n\n")


new_state2 = graph.get_state(config)
print("graph_state.value =", new_state2.values)

Error Message and Stack Trace (if applicable)

Description

I work with graphs and subgraphs that can be interrupted. When I update the state of the parent graph after interrupting a node in the subgraph, two issues occur:

  • The parent graph’s state doesn’t reflect the latest execution result from the subgraph’s node.
  • The subgraph restarts from the beginning instead of resuming from where it was interrupted.

I expect to see:

  1. Executing node1
  2. Executing subgraph_node_1
  3. [Updating the graph state]
  4. Executing subgraph_node_2
  5. graph_state.value = {'foo': True, 'bar': True}

Instead, it does:

  1. Executing node1
  2. Executing subgraph_node_1
  3. [Updating the graph state]
  4. Executing subgraph_node_1
  5. graph_state.value = {'foo': False, 'bar': True}

Also I drew the graph as a PNG image using a mermaid:

Image

System Info

System Information

OS: Linux
OS Version: #61~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 15 17:03:15 UTC 2
Python Version: 3.12.7 (main, Oct 10 2024, 15:12:39) [GCC 9.4.0]

Package Information

langchain_core: 0.3.60
langsmith: 0.3.42
langgraph_sdk: 0.1.69

Optional packages not installed

langserve

Other Dependencies

httpx: 0.28.1
jsonpatch<2.0,>=1.33: Installed. No version info available.
langsmith-pyo3: Installed. No version info available.
langsmith<0.4,>=0.1.126: Installed. No version info available.
openai-agents: Installed. No version info available.
opentelemetry-api: Installed. No version info available.
opentelemetry-exporter-otlp-proto-http: Installed. No version info available.
opentelemetry-sdk: Installed. No version info available.
orjson: 3.10.18
packaging: 24.2
packaging<25,>=23.2: Installed. No version info available.
pydantic: 2.11.4
pydantic>=2.7.4: Installed. No version info available.
pytest: Installed. No version info available.
PyYAML>=5.3: Installed. No version info available.
requests: 2.32.3
requests-toolbelt: 1.0.0
rich: Installed. No version info available.
tenacity!=8.4.0,<10.0.0,>=8.1.0: Installed. No version info available.
typing-extensions>=4.7: Installed. No version info available.
zstandard: 0.23.0

Originally created by @egortaran on GitHub (May 19, 2025). ### Checked other resources - [x] This is a bug, not a usage question. For questions, please use GitHub Discussions. - [x] I added a clear and detailed title that summarizes the issue. - [x] I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example). - [x] I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue. ### Example Code ```python from typing import TypedDict from langchain_core.runnables import RunnableConfig from langgraph.checkpoint.memory import MemorySaver from langgraph.constants import START from langgraph.graph import END, StateGraph class State(TypedDict): foo: bool bar: bool # Preparing subgraph def subgraph_node_1(state: State) -> State: state["foo"] = True print("Executing `subgraph_node_1`") return state def subgraph_node_2(state: State) -> State: print("Executing `subgraph_node_2`") return state subgraph_builder = StateGraph(State) subgraph_builder.add_node("subgraph_node_1", subgraph_node_1) subgraph_builder.add_node("subgraph_node_2", subgraph_node_2) subgraph_builder.add_edge(START, "subgraph_node_1") subgraph_builder.add_edge("subgraph_node_1", "subgraph_node_2") subgraph_builder.add_edge("subgraph_node_2", END) subgraph = subgraph_builder.compile( interrupt_after=[ "subgraph_node_1", "subgraph_node_2", ], ) # Preparing parent graph builder = StateGraph(State) def node1(state: State) -> State: print("Executing `node1`") return state def node3(state: State) -> State: print("Executing `node3`") return state builder.add_node("node1", node1) builder.add_node("node2", subgraph) builder.add_node("node3", node3) builder.add_edge(START, "node1") builder.add_edge("node1", "node2") builder.add_edge("node2", "node3") builder.add_edge("node3", END) # Graph executing checkpointer = MemorySaver() config = RunnableConfig(configurable={"thread_id": "1"}) graph = builder.compile( checkpointer=checkpointer, interrupt_after=["node1", "node3",], ) for event in graph.stream({"foo": False, "bar": False}, config=config, stream_mode="updates", subgraphs=True): print(event) print("*" * 10, "\n\n") for event in graph.stream(None, config, subgraphs=True): print(event) print("*" * 10, "\n\n") # Update graph state new_config = graph.update_state(config, {"bar": True}) for event in graph.stream(None, new_config, subgraphs=True): print(event) print("*" * 10, "\n\n") new_state2 = graph.get_state(config) print("graph_state.value =", new_state2.values) ``` ### Error Message and Stack Trace (if applicable) ```shell ``` ### Description I work with graphs and subgraphs that can be interrupted. When I update the state of the parent graph after interrupting a node in the subgraph, two issues occur: - The parent graph’s state doesn’t reflect the latest execution result from the subgraph’s node. - The subgraph restarts from the beginning instead of resuming from where it was interrupted. I expect to see: 1. Executing node1 2. Executing subgraph_node_1 3. [Updating the graph state] 4. Executing **subgraph_node_2** 5. graph_state.value = {'foo': **True**, 'bar': True} Instead, it does: 1. Executing node1 2. Executing subgraph_node_1 3. [Updating the graph state] 4. Executing **subgraph_node_1** 5. graph_state.value = {'foo': **False**, 'bar': True} Also I drew the graph as a PNG image using a mermaid: ![Image](https://github.com/user-attachments/assets/214ef4e9-d289-4d1b-85db-efbb1064bc35) ### System Info System Information ------------------ > OS: Linux > OS Version: #61~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Tue Apr 15 17:03:15 UTC 2 > Python Version: 3.12.7 (main, Oct 10 2024, 15:12:39) [GCC 9.4.0] Package Information ------------------- > langchain_core: 0.3.60 > langsmith: 0.3.42 > langgraph_sdk: 0.1.69 Optional packages not installed ------------------------------- > langserve Other Dependencies ------------------ > httpx: 0.28.1 > jsonpatch<2.0,>=1.33: Installed. No version info available. > langsmith-pyo3: Installed. No version info available. > langsmith<0.4,>=0.1.126: Installed. No version info available. > openai-agents: Installed. No version info available. > opentelemetry-api: Installed. No version info available. > opentelemetry-exporter-otlp-proto-http: Installed. No version info available. > opentelemetry-sdk: Installed. No version info available. > orjson: 3.10.18 > packaging: 24.2 > packaging<25,>=23.2: Installed. No version info available. > pydantic: 2.11.4 > pydantic>=2.7.4: Installed. No version info available. > pytest: Installed. No version info available. > PyYAML>=5.3: Installed. No version info available. > requests: 2.32.3 > requests-toolbelt: 1.0.0 > rich: Installed. No version info available. > tenacity!=8.4.0,<10.0.0,>=8.1.0: Installed. No version info available. > typing-extensions>=4.7: Installed. No version info available. > zstandard: 0.23.0
yindo closed this issue 2026-02-20 17:41:02 -05:00
yindo commented 2026-02-20 17:41:02 -05:00
Author
Owner
Copy Link
Copy Source

@egortaran commented on GitHub (May 19, 2025):

I’m not sure why the LangGraph version isn’t displayed when I run python -m langchain_core.sys_info, so I’m providing the version details manually pip show langgraph:

Name: langgraph
Version: 0.4.5
Summary: Building stateful, multi-actor applications with LLMs
Home-page: 
Author: 
Author-email: 
License: MIT
Location: /home/taran/PycharmProjects/PythonProject/.venv/lib/python3.12/site-packages
Requires: langchain-core, langgraph-checkpoint, langgraph-prebuilt, langgraph-sdk, pydantic, xxhash
Required-by:
@egortaran commented on GitHub (May 19, 2025): I’m not sure why the LangGraph version isn’t displayed when I run `python -m langchain_core.sys_info`, so I’m providing the version details manually `pip show langgraph`: ``` Name: langgraph Version: 0.4.5 Summary: Building stateful, multi-actor applications with LLMs Home-page: Author: Author-email: License: MIT Location: /home/taran/PycharmProjects/PythonProject/.venv/lib/python3.12/site-packages Requires: langchain-core, langgraph-checkpoint, langgraph-prebuilt, langgraph-sdk, pydantic, xxhash Required-by: ```
yindo commented 2026-02-20 17:41:02 -05:00
Author
Owner
Copy Link
Copy Source

@michalgala commented on GitHub (May 25, 2025):

try adding checkpointer=True in the subgraph compilation

subgraph = subgraph_builder.compile(
    checkpointer=True,
    interrupt_after=[
        "subgraph_node_1",
        "subgraph_node_2",
    ],
)
@michalgala commented on GitHub (May 25, 2025): try adding _checkpointer=True_ in the subgraph compilation ``` subgraph = subgraph_builder.compile( checkpointer=True, interrupt_after=[ "subgraph_node_1", "subgraph_node_2", ], ) ```
yindo commented 2026-02-20 17:41:02 -05:00
Author
Owner
Copy Link
Copy Source

@egortaran commented on GitHub (May 27, 2025):

Thanks for the answer, @michalgala! Your comment solved the problem in half, which is already good!
The issue persists when retrieving the graph state values. That is, if I apply your changes, my log will be:

Executing `node1`
((), {'node1': {'foo': False, 'bar': False}})
((), {'__interrupt__': ()})
********** 


Executing `subgraph_node_1`
(('node2',), {'subgraph_node_1': {'foo': True, 'bar': False}})
((), {'__interrupt__': ()})
********** 


Executing `subgraph_node_2`
(('node2',), {'subgraph_node_2': {'foo': True, 'bar': False}})
((), {'__interrupt__': ()})
********** 


graph_state.value = {'foo': False, 'bar': True}

In the "Executing subgraph_node_2" I expected to see:

{'subgraph_node_2': {'foo': True, 'bar': True}}

Instead, it does:

{'subgraph_node_2': {'foo': True, 'bar': False}}

Also in the last print I expected to see:

graph_state.value = {'foo': True, 'bar': True}

Instead, it does:

graph_state.value = {'foo': False, 'bar': True}
@egortaran commented on GitHub (May 27, 2025): Thanks for the answer, @michalgala! Your comment solved the problem in half, which is already good! The issue persists when retrieving the graph state values. That is, if I apply your changes, my log will be: ``` Executing `node1` ((), {'node1': {'foo': False, 'bar': False}}) ((), {'__interrupt__': ()}) ********** Executing `subgraph_node_1` (('node2',), {'subgraph_node_1': {'foo': True, 'bar': False}}) ((), {'__interrupt__': ()}) ********** Executing `subgraph_node_2` (('node2',), {'subgraph_node_2': {'foo': True, 'bar': False}}) ((), {'__interrupt__': ()}) ********** graph_state.value = {'foo': False, 'bar': True} ``` ___ In the "**Executing subgraph_node_2**" I expected to see: ``` {'subgraph_node_2': {'foo': True, 'bar': True}} ``` Instead, it does: ``` {'subgraph_node_2': {'foo': True, 'bar': False}} ``` ___ Also in the last print I expected to see: ``` graph_state.value = {'foo': True, 'bar': True} ``` Instead, it does: ``` graph_state.value = {'foo': False, 'bar': True} ```
yindo commented 2026-02-20 17:41:03 -05:00
Author
Owner
Copy Link
Copy Source

@michalgala commented on GitHub (May 27, 2025):

Yes, now we have the same issue, see here

There is a workaround, you could use interrupt(), and resume your graph with Command(resume=..)

@michalgala commented on GitHub (May 27, 2025): Yes, now we have the same issue, [see here](https://github.com/langchain-ai/langgraph/issues/4796) There is a workaround, you could use interrupt(), and resume your graph with Command(resume=..)
yindo commented 2026-02-20 17:41:03 -05:00
Author
Owner
Copy Link
Copy Source

@sydney-runkle commented on GitHub (Jun 11, 2025):

Closing this as a dupe of https://github.com/langchain-ai/langgraph/issues/4866, we'll provide updates there!

@sydney-runkle commented on GitHub (Jun 11, 2025): Closing this as a dupe of https://github.com/langchain-ai/langgraph/issues/4866, we'll provide updates there!
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#633
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?