Unable to work properly with SSE transport when perform long time task

[huang-sh]

Dear mcp team,

Describe the bug
I'm building an mcp-server that reads large files, which can be time-consuming.

When using stdio as the transport, everything works well. However, when switching to sse, the server fails to work properly. Could you provide some help?

WARNING:root:Failed to validate notification: RequestResponder must be used as a context manager. 
Message was: method='notifications/cancelled' params={'requestId': 2, 'reason': 'McpError: MCP error -32001: Request timed out'} jsonrpc='2.0'

{
    "isError": true,
    "content": [
      {
        "type": "text",
        "text": "Error calling tool read_h5ad: {}"
      }
    ]
  }

sse start code

## Create application using SSE transport
def create_sse_app(port=8000):
    """
    Create application using SSE transport
    
    Parameters:
        port: Server port number
        
    Returns:
        Starlette application instance
    """
    from starlette.applications import Starlette
    from starlette.routing import Route, Mount
    from starlette.requests import Request
    from mcp.server.sse import SseServerTransport
    
    # Create SSE transport object
    sse = SseServerTransport("/messages/")

    # Define SSE handler function
    async def handle_sse(request):
        async with sse.connect_sse(
            request.scope, request.receive, request._send
        ) as streams:
            await server.run(
                streams[0], streams[1], 
                InitializationOptions(
                    server_name=f"scanpy-mcp-{MODULE}",
                    server_version="0.1.2",
                    capabilities=server.get_capabilities(
                        notification_options=NotificationOptions(),
                        experimental_capabilities={},
                    ),
                )
            )

    # Create Starlette application
    starlette_app = Starlette(
        routes=[
            Route("/sse", endpoint=handle_sse),
            Mount("/messages/", app=sse.handle_post_message),
            Route("/figures/{figure_name}", endpoint=get_figure),
        ]
    )
    
    return starlette_app

import uvicorn
app = create_sse_app(port=port)
uvicorn.run(app, host="0.0.0.0", port=port)

[aaadrain]

Hi，how to solve it .

[felixweinberger]

Hi @huang-sh thanks for this submission.

How long are the long running tasks running for you? We have some default timeouts that you might be running into (e.g. if these tasks run much longer than 5 minutes).

If you could provide a code snippet that reproduces the problem you're seeing that'd be great so we can debug on our machines.

[maxisbey]

Closing as stale — we asked for additional information and haven't heard back. If you're still experiencing this on the latest release, please reopen with a minimal reproducible example and we'll take another look.

_{AI Disclaimer}