ClientSession for mcp sse_client throws httpx.ReadTimeout after session has been initialized in docker container

[cmakasy]

Initial Checks

• [x] I confirm that I'm using the latest version of MCP Python SDK
• [x] I confirm that I searched for my issue in https://github.com/modelcontextprotocol/python-sdk/issues before opening this issue

Description

We are implementing a FastAPI mcp client that uses an external sse server (http://localhost:9000/sse). Form this code we created a dockerfile that runs into an httpx.ReadTimeout Exception Running the client from the docker image shows the following output:

      INFO   Started server process [276]
      INFO   Waiting for application startup.
🌀 Tâche d'initialisation MCP lancée en arrière-plan.
🚀 Initialisation de la session MCP...
      INFO   Application startup complete.
SSE CLIENT passed
 initialized – protocol version: meta=None protocolVersion='2024-11-05' capabilities=ServerCapabilities(experimental=None, logging=LoggingCapability(), prompts=PromptsCapability(listChanged=True), resources=ResourcesCapability(subscribe=False, listChanged=True), tools=ToolsCapability(listChanged=True), completions=CompletionsCapability()) serverInfo=Implementation(name='papylot-mcp-server', title=None, version='0.0.1', websiteUrl=None, icons=None) instructions=None
✅ MCP session initialized.
PING meta=None

After a couple of time a httpx.ReadTimeout will be shown.

Full exception:

Traceback (most recent call last):
  File "/usr/local/lib/python3.12/site-packages/mcp/client/sse.py", line 73, in sse_reader
    async for sse in event_source.aiter_sse():
  File "/usr/local/lib/python3.12/site-packages/httpx_sse/_api.py", line 39, in aiter_sse
    async for line in self._response.aiter_lines():
  File "/usr/local/lib/python3.12/site-packages/httpx/_models.py", line 1031, in aiter_lines
    async for text in self.aiter_text():
  File "/usr/local/lib/python3.12/site-packages/httpx/_models.py", line 1018, in aiter_text
    async for byte_content in self.aiter_bytes():
  File "/usr/local/lib/python3.12/site-packages/httpx/_models.py", line 997, in aiter_bytes
    async for raw_bytes in self.aiter_raw():
  File "/usr/local/lib/python3.12/site-packages/httpx/_models.py", line 1055, in aiter_raw
    async for raw_stream_bytes in self.stream:
  File "/usr/local/lib/python3.12/site-packages/httpx/_client.py", line 176, in __aiter__
    async for chunk in self._stream:
  File "/usr/local/lib/python3.12/site-packages/httpx/_transports/default.py", line 270, in __aiter__
    with map_httpcore_exceptions():
         ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/contextlib.py", line 158, in __exit__
    self.gen.throw(value)
  File "/usr/local/lib/python3.12/site-packages/httpx/_transports/default.py", line 118, in map_httpcore_exceptions
    raise mapped_exc(message) from exc
httpx.ReadTimeout

Example Code

import os
import re
import json
import asyncio
import httpx
from fastapi import FastAPI, Query
from dotenv import load_dotenv
from mcp import ClientSession
from pydantic_ai import Agent
# from groq import Groq
from fastmcp import FastMCP
from contextlib import asynccontextmanager
from mcp.client.sse import sse_client

async def init_mcp_background():
      async with  sse_client("http://localhost:9000/sse") as (read, write):
            async with ClientSession(read, write) as sess:
                session = sess
                await session.initialize()
                print("✅ MCP session initialized.")

                ping = await session.send_ping()
                print("ping",ping)
                
                #print("info",info)
                tools = await session.list_tools()
                print("Available TOOLS", tools.tools)

                # 🔁 Boucle qui garde la connexion active
                while True:
                    await asyncio.sleep(10) 

@asynccontextmanager
async def lifespan(app: FastAPI):
    asyncio.create_task(init_mcp_background())
    print("🌀 Tâche d'initialisation MCP lancée en arrière-plan.")
    yield

app = FastAPI(lifespan=lifespan)
    @app.get("/query")
async def get_response(question: str = Query(..., description="La question de l'utilisateur")):
    
    return{"question", question}


with the dockerfile for image creation:

# syntax=docker/dockerfile:1
## DOCKERFILE to containerize BACKEND ##
FROM python:3.12-slim
WORKDIR /app
# copy requirements
COPY ./requirements.txt /app/requirements.txt
# install requirements
RUN pip install --no-cache-dir --upgrade -r /app/requirements.txt
# copy source code
COPY .  /app
# where are we
RUN pwd
# what do we have
RUN ls -lR
# run application
CMD ["fastapi", "run", "API/api.py", "--proxy-headers", "--port", "8011"]

Python & MCP Python SDK

For the dockerfile we are using: :3.12-slim
and the requirement.txt is

fastmcp==2.12.4
mcp==1.16.0
pydantic_ai==1.0.14
python-dotenv==1.1.1
fastapi-mcp==0.3.4
fastapi[standard]>=0.115.12