Fix MCP client HTTP transport compatibility for multiple calls #852
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Fix 'Task group is not initialized' error on subsequent MCP calls - Implement persistent event loop management in MCP data collector - Update to non-deprecated streamable_http_client API - Add support for Streamlit environment with thread pool execution - Upgrade MCP SDK packages (fastmcp>=2.14.1, mcp>=1.25.0) - Configure httpx timeout for long-running operations - Rename sse_read_timeout to read_timeout for clarity Changes: - mcp_server/requirements.txt: Upgrade fastmcp and mcp packages - mcp_server/server.py: Use streamable-http transport - mcp_server/run_server.sh: Update transport configuration - tools/mcp_data_collector.py: Implement persistent event loop and retry logic - tools/release_progress_dashboard/requirements.txt: Update dependencies Root cause: Using asyncio.run() created new event loop per call, breaking FastMCP session management. Fixed by maintaining persistent event loop and using loop.run_until_complete() for reusable sessions.
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
@rioliu-rh: all tests passed! Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
rioliu-rh
added a commit
to rioliu-rh/release-tests
that referenced
this pull request
Dec 23, 2025
- Update architecture diagram from HTTP/SSE to HTTP - Update server transport from sse to streamable-http - Update Claude Code client config from transport:sse to type:http - Update Python client examples to use streamable_http_client - Update all /sse endpoints to /mcp - Update server configuration examples with new host/port API - Update curl testing commands This completes the documentation updates for the HTTP transport migration started in PR openshift#852.
rioliu-rh
added a commit
that referenced
this pull request
Dec 23, 2025
* Update MCP deployment guide for HTTP transport migration - Update architecture diagram from HTTP/SSE to HTTP - Update server transport from sse to streamable-http - Update Claude Code client config from transport:sse to type:http - Update Python client examples to use streamable_http_client - Update all /sse endpoints to /mcp - Update server configuration examples with new host/port API - Update curl testing commands This completes the documentation updates for the HTTP transport migration started in PR #852. * Update dashboard deployment guide for HTTP transport migration - Update all MCP server URLs from /sse to /mcp endpoint - Update environment variable examples - Update systemd service configuration examples - Update troubleshooting curl commands This ensures dashboard deployment documentation is consistent with the MCP server HTTP transport migration.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
Fixes the "Task group is not initialized" error that occurs on subsequent MCP client calls when using HTTP transport. The first call succeeds but subsequent calls fail with 500 Internal Server Error.
Related: OCPERT-275
Root Cause
The MCP data collector was using
asyncio.run()for each call, which creates a new event loop every time. FastMCP's streamable-http transport requires persistent event loop and session management - creating new loops breaks the internal task group state.This is a known issue in the MCP ecosystem:
Changes
MCP Server:
MCP Data Collector Client:
streamable_http_clientAPITesting
Verified with test script making 3 consecutive calls:
Deployment Notes
After merge, remote server needs:
pip3 install -r mcp_server/requirements.txt --upgradecurl http://localhost:8000/healthCompatibility