🐛 Bug Report: Resetting async context doesn't work after update to 0.44.0

[zmievsa-semasoftware]

Which component is this bug for?

Langchain Instrumentation

📜 Description

Ever since I updated from 0.40.14, I started getting the following errors. It seems like it happens because I am jumping from sync to async and back but I am not very sure.

Errors

This one starts at 0.44.0:

Traceback (most recent call last):
  File "<project_root>/.venv/lib/python3.12/site-packages/opentelemetry/context/__init__.py", line 155, in detach
    _RUNTIME_CONTEXT.detach(token)
  File "<project_root>/.venv/lib/python3.12/site-packages/opentelemetry/context/contextvars_context.py", line 53, in detach
    self._current_context.reset(token)
ValueError: <Token var=<ContextVar name='current_context' default={} at 0x103000090> at 0x15fc8a200> was created in a different Context

And looks like this is the offender: 168236a

Opentelemetry lib versions

opentelemetry-api==1.36.0
opentelemetry-distro==0.57b0
opentelemetry-exporter-otlp==1.36.0
opentelemetry-exporter-otlp-proto-common==1.36.0
opentelemetry-exporter-otlp-proto-grpc==1.36.0
opentelemetry-exporter-otlp-proto-http==1.36.0
opentelemetry-instrumentation==0.57b0
opentelemetry-instrumentation-aiohttp-client==0.57b0
opentelemetry-instrumentation-aiohttp-server==0.57b0
opentelemetry-instrumentation-asgi==0.57b0
opentelemetry-instrumentation-asyncio==0.57b0
opentelemetry-instrumentation-boto3sqs==0.57b0
opentelemetry-instrumentation-botocore==0.57b0
opentelemetry-instrumentation-click==0.57b0
opentelemetry-instrumentation-dbapi==0.57b0
opentelemetry-instrumentation-django==0.57b0
opentelemetry-instrumentation-grpc==0.57b0
opentelemetry-instrumentation-httpx==0.57b0
opentelemetry-instrumentation-jinja2==0.57b0
opentelemetry-instrumentation-langchain==0.44.1
opentelemetry-instrumentation-logging==0.57b0
opentelemetry-instrumentation-openai-v2==2.1b0
opentelemetry-instrumentation-psycopg==0.57b0
opentelemetry-instrumentation-requests==0.57b0
opentelemetry-instrumentation-sqlalchemy==0.57b0
opentelemetry-instrumentation-sqlite3==0.57b0
opentelemetry-instrumentation-threading==0.57b0
opentelemetry-instrumentation-tortoiseorm==0.57b0
opentelemetry-instrumentation-urllib==0.57b0
opentelemetry-instrumentation-urllib3==0.57b0
opentelemetry-instrumentation-wsgi==0.57b0
opentelemetry-propagator-aws-xray==1.0.2
opentelemetry-proto==1.36.0
opentelemetry-sdk==1.36.0
opentelemetry-semantic-conventions==0.57b0
opentelemetry-semantic-conventions-ai==0.4.11
opentelemetry-util-http==0.57b0

👟 Reproduction steps

Still working on it, it's a part of the big project so I'm still wondering what causes it. But it is definitely related to update of opentelemetry-instrumentation-langchain from 0.40.14.

👍 Expected behavior

It should produce no such errors as it did before

👎 Actual Behavior with Screenshots

It produces a lot of ValueErrors when trying to reset context.

🤖 Python Version

3.12.10

📃 Provide any additional context for the Bug.

No response

👀 Have you spent some time to check if this bug has been raised before?

• I checked and didn't find similar issue

Are you willing to submit PR?

None

[gcctwo]

I'm experiencing the same issue. It shouldn't be hard to reproduce, but here is a simple example that produces the bug:

from dotenv import load_dotenv
from langchain_aws import ChatBedrock
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser
from opentelemetry.instrumentation.langchain import LangchainInstrumentor

load_dotenv()
LangchainInstrumentor().instrument()

llm = ChatBedrock(
    region_name="us-east-1",
    model_id="anthropic.claude-3-sonnet-20240229-v1:0",
)

async def main():
    query = "What are the benefits of renewable energy?"
    try:
        prompt = ChatPromptTemplate.from_messages([
            ("system", "Be extremely concise. Simple and short answers are preferred."),
            ("human", "Query: {query}")
        ])
        chain = prompt | llm | StrOutputParser()
        result = await chain.ainvoke({"query": query})
        print(f"Final Result: {result[:200]}...")
    except Exception as e:
        print(f"Error: {e}")

if __name__ == "__main__":
    asyncio.run(main())

[zmievsa]

Wait, it might've been fixed here #3255