mirrors/hermes-agent
1
0
Fork 0
mirror of https://github.com/NousResearch/hermes-agent.git synced 2026-07-22 16:25:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

fix(mcp): allow background discovery retry after a run that connected nothing

Browse source 
start_background_mcp_discovery() sets _mcp_discovery_started once and never
resets it. If the first background run exits without connecting any MCP
server (startup cancellation, OOM restart, transient network failure), every
later call returns immediately and the process is permanently stuck with
zero MCP tools until a full restart.

Fix: when discovery is marked started but the thread is dead and no server
is connected, reset the flag and spawn a fresh discovery thread. Also log a
WARNING when a discovery run completes with zero connected servers, so the
condition is visible instead of silent.

Caught in production on a long-running gateway fleet where a gateway
restarted under memory pressure and came back with all MCP tools missing.
This commit is contained in:
fazerluga-creator 2026-07-18 17:55:02 +03:00 • committed by Teknium
parent fbe086f7cc
commit 4fda7efcbd
2 changed files with 111 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

40
hermes_cli/mcp_startup.py
View file

@ -25,12 +25,35 @@ def _has_configured_mcp_servers() -> bool:
def start_background_mcp_discovery(*, logger, thread_name: str) -> None:
"""Spawn one shared background MCP discovery thread for this process."""
"""Spawn one shared background MCP discovery thread for this process.
If the first background discovery run exits without connecting any MCP
server (for example after startup cancellation / OOM restart), later calls
are allowed to retry instead of permanently pinning the process in a
"discovery already started" state with zero MCP tools.
"""
global _mcp_discovery_started, _mcp_discovery_thread
with _mcp_discovery_lock:
if _mcp_discovery_started:
return
thread = _mcp_discovery_thread
if thread is not None and thread.is_alive():
return
try:
from tools.mcp_tool import get_mcp_status
status = get_mcp_status() or []
if any(entry.get("connected") for entry in status):
return
except Exception:
return
logger.warning(
"Background MCP discovery previously exited with no connected "
"servers; retrying discovery thread"
)
_mcp_discovery_started = False
_mcp_discovery_thread = None
_mcp_discovery_started = True
if not _has_configured_mcp_servers():
return
@ -38,8 +61,21 @@ def start_background_mcp_discovery(*, logger, thread_name: str) -> None:
def _discover() -> None:
try:
_discover_mcp_tools_without_interactive_oauth()
try:
from tools.mcp_tool import get_mcp_status
status = get_mcp_status() or []
if not any(entry.get("connected") for entry in status):
logger.warning(
"Background MCP discovery completed with zero connected servers"
)
except Exception:
logger.debug("Failed to inspect MCP status after background discovery", exc_info=True)
except Exception:
logger.debug("Background MCP tool discovery failed", exc_info=True)
finally:
with _mcp_discovery_lock:
global _mcp_discovery_thread, _mcp_discovery_started
_mcp_discovery_thread = None
thread = threading.Thread(
target=_discover,