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

[PR #6542] fix(langgraph): fix interrupt resume routing for parallel interrupts #5112

New Issue
Closed
opened 2026-02-20 17:51:14 -05:00 by yindo · 0 comments
yindo commented 2026-02-20 17:51:14 -05:00
Owner
Copy Link
Copy Source

Original Pull Request: https://github.com/langchain-ai/langgraph/pull/6542

State: closed
Merged: No

Summary

Fix interrupt resume routing for parallel interrupts (#6533, #6534).

When multiple interrupt() calls occur within the same task (e.g., parallel tools in ToolNode), they were generating identical interrupt IDs because the ID was derived solely from the checkpoint namespace. This caused resume values to be misrouted between tools.

Problem

When using ToolNode with multiple tools that call interrupt():

@tool
def tool_a(val: str) -> str:
    response = interrupt({"tool": "a"})  # ID: abc123
    return f"A got: {response}"

@tool  
def tool_b(val: str) -> str:
    response = interrupt({"tool": "b"})  # ID: abc123 (SAME!)
    return f"B got: {response}"

Both tools shared the same checkpoint namespace, so both interrupts received identical IDs. When resuming with Command(resume={id: value}), values could be routed to the wrong tool.

Solution

  1. Include interrupt counter in ID generation (types.py):

    # Before
ns=conf[CONFIG_KEY_CHECKPOINT_NS]
# After  
ns=f"{conf[CONFIG_KEY_CHECKPOINT_NS]}|{idx}"

  2. Update resume_map lookup (_algo.py):

    • Changed _scratchpad() to iterate through counter indices when looking up resume values
    • Ensures all interrupt-specific resume values are found

Testing

  • Added test_tool_node_parallel_interrupts in prebuilt (regression test for #6533)
  • Added test_interrupt_resume_with_multiple_nodes in langgraph (regression test for #6534)
  • All 189 existing interrupt tests pass

Fixes #6533
Fixes #6534

**Original Pull Request:** https://github.com/langchain-ai/langgraph/pull/6542 **State:** closed **Merged:** No --- ## Summary Fix interrupt resume routing for parallel interrupts (#6533, #6534). When multiple `interrupt()` calls occur within the same task (e.g., parallel tools in ToolNode), they were generating identical interrupt IDs because the ID was derived solely from the checkpoint namespace. This caused resume values to be misrouted between tools. ## Problem When using ToolNode with multiple tools that call `interrupt()`: ```python @tool def tool_a(val: str) -> str: response = interrupt({"tool": "a"}) # ID: abc123 return f"A got: {response}" @tool def tool_b(val: str) -> str: response = interrupt({"tool": "b"}) # ID: abc123 (SAME!) return f"B got: {response}" ``` Both tools shared the same checkpoint namespace, so both interrupts received identical IDs. When resuming with `Command(resume={id: value})`, values could be routed to the wrong tool. ## Solution 1. **Include interrupt counter in ID generation** (`types.py`): ```python # Before ns=conf[CONFIG_KEY_CHECKPOINT_NS] # After ns=f"{conf[CONFIG_KEY_CHECKPOINT_NS]}|{idx}" ``` 2. **Update resume_map lookup** (`_algo.py`): - Changed `_scratchpad()` to iterate through counter indices when looking up resume values - Ensures all interrupt-specific resume values are found ## Testing - Added `test_tool_node_parallel_interrupts` in prebuilt (regression test for #6533) - Added `test_interrupt_resume_with_multiple_nodes` in langgraph (regression test for #6534) - All 189 existing interrupt tests pass Fixes #6533 Fixes #6534
yindo added the pull-request label 2026-02-20 17:51:14 -05:00
yindo closed this issue 2026-02-20 17:51:14 -05:00
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 pull-request
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#5112
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?