From 60674e326f36913213ac430f31e8a7c522efb522 Mon Sep 17 00:00:00 2001
From: liuhao1024 <sunsky.lau@gmail.com>
Date: Sat, 30 May 2026 18:07:42 +0800
Subject: [PATCH] fix(telegram): clear send_path_degraded immediately on
 successful reconnect

After a network error, _send_path_degraded was only cleared by the
60-second heartbeat probe in _verify_polling_after_reconnect(). During
that window, all outbound send() calls short-circuited as failures,
even though polling had already resumed successfully.

Clear the flag immediately after start_polling() succeeds. If the
send-side httpx pool is also stale, send()'s own retry and
error_callback will re-enter the reconnect ladder.

Fixes #35205
---
 gateway/platforms/telegram.py                 |  6 ++
 .../gateway/test_telegram_send_path_health.py | 55 +++++++++++++++++--
 2 files changed, 56 insertions(+), 5 deletions(-)

diff --git a/gateway/platforms/telegram.py b/gateway/platforms/telegram.py
index 7b4d00e818fc..10aca46118ef 100644
--- a/gateway/platforms/telegram.py
+++ b/gateway/platforms/telegram.py
@@ -962,6 +962,12 @@ async def _handle_polling_network_error(self, error: Exception) -> None:
                 self.name, attempt,
             )
             self._polling_network_error_count = 0
+            # Polling is confirmed working — clear the degraded flag so
+            # send() can attempt delivery immediately instead of waiting
+            # 60 s for the heartbeat probe.  If the send-side httpx pool
+            # is also stale, send()'s own retry + error_callback will
+            # re-enter the reconnect ladder.
+            self._send_path_degraded = False
             # start_polling() returning is necessary but not sufficient:
             # PTB's Updater can be left in a state where `running` is True
             # but the underlying long-poll task is wedged on a stale httpx
diff --git a/tests/gateway/test_telegram_send_path_health.py b/tests/gateway/test_telegram_send_path_health.py
index 05972bdba437..77545d340f54 100644
--- a/tests/gateway/test_telegram_send_path_health.py
+++ b/tests/gateway/test_telegram_send_path_health.py
@@ -5,6 +5,7 @@
 but nothing reaches the recipient.  ``_send_path_degraded`` short-circuits
 ``send()`` so cron's live-adapter branch falls through to standalone HTTP.
 """
+import asyncio
 import sys
 from unittest.mock import AsyncMock, MagicMock, patch
 
@@ -65,15 +66,19 @@ async def test_send_short_circuits_when_path_degraded():
 
 
 @pytest.mark.asyncio
-async def test_reconnect_storm_sets_and_heartbeat_clears_flag(monkeypatch):
-    """_handle_polling_network_error sets the flag; a successful heartbeat
-    probe in _verify_polling_after_reconnect clears it."""
+async def test_reconnect_clears_degraded_flag_on_successful_start_polling(monkeypatch):
+    """After a successful start_polling() reconnect, _send_path_degraded must
+    be cleared immediately — not deferred to the 60-second heartbeat probe.
+
+    Regression test for #35205: send_path_degraded stayed True for 60 s after
+    reconnect, blocking all outbound messages.
+    """
     adapter = _make_adapter()
     adapter._app = MagicMock()
     adapter._app.updater = MagicMock()
     adapter._app.updater.running = True
     adapter._app.updater.stop = AsyncMock()
-    adapter._app.updater.start_polling = AsyncMock()
+    adapter._app.updater.start_polling = AsyncMock()  # succeeds
     adapter._app.bot = MagicMock()
     adapter._app.bot.get_me = AsyncMock(return_value=MagicMock())
     adapter._polling_error_callback_ref = AsyncMock()
@@ -82,8 +87,48 @@ async def test_reconnect_storm_sets_and_heartbeat_clears_flag(monkeypatch):
     )
 
     await adapter._handle_polling_network_error(OSError("Bad Gateway"))
-    assert adapter._send_path_degraded is True
+    # Flag must be cleared immediately after successful start_polling()
+    assert adapter._send_path_degraded is False
 
+    # Heartbeat probe still runs and confirms the flag stays cleared
     with patch("gateway.platforms.telegram.asyncio.sleep", new_callable=AsyncMock):
         await adapter._verify_polling_after_reconnect()
     assert adapter._send_path_degraded is False
+
+
+@pytest.mark.asyncio
+async def test_degraded_flag_stays_true_when_start_polling_fails(monkeypatch):
+    """When start_polling() fails, _send_path_degraded must remain True
+    so send() continues to short-circuit until recovery succeeds."""
+    adapter = _make_adapter()
+    adapter._polling_network_error_count = 1
+
+    mock_updater = MagicMock()
+    mock_updater.running = True
+    mock_updater.stop = AsyncMock()
+    mock_updater.start_polling = AsyncMock(side_effect=Exception("Timed out"))
+
+    mock_app = MagicMock()
+    mock_app.updater = mock_updater
+    adapter._app = mock_app
+    adapter._polling_error_callback_ref = AsyncMock()
+    monkeypatch.setattr(
+        "gateway.platforms.telegram.Update", MagicMock(ALL_TYPES=[])
+    )
+
+    with patch("asyncio.sleep", new_callable=AsyncMock):
+        await adapter._handle_polling_network_error(Exception("Bad Gateway"))
+
+    # Flag must stay True since reconnect failed
+    assert adapter._send_path_degraded is True
+
+    # Clean up pending retry tasks
+    for t in list(adapter._background_tasks):
+        t.cancel()
+        try:
+            await t
+        except (asyncio.CancelledError, Exception):
+            pass
+
+
+