From 1a93caf62de3ebe66fee678b17f38f89d21a652b Mon Sep 17 00:00:00 2001
From: "Gan, Jimmy" <jimmy.gan@adidas.com>
Date: Fri, 20 Feb 2026 10:31:18 +0800
Subject: [PATCH] feat: add container monitor & telegram alerts

Added asyncio monitoring task to main.py to watch Docker container state transitions and send Telegram notifications on failures.
---
 dashboard/backend/config.py         |  4 ++++
 dashboard/backend/main.py           | 37 +++++++++++++++++++++++++++++
 dashboard/backend/routers/system.py |  4 +---
 3 files changed, 42 insertions(+), 3 deletions(-)

diff --git a/dashboard/backend/config.py b/dashboard/backend/config.py
index 61c98dc..82f629b 100644
--- a/dashboard/backend/config.py
+++ b/dashboard/backend/config.py
@@ -23,3 +23,7 @@ SSH_KEY_PATH = os.environ.get("SSH_KEY_PATH", "/app/ssh/id_ed25519")
 VPS_SSH_HOST = os.environ.get("VPS_SSH_HOST", "158.101.140.85")
 VPS_SSH_USER = os.environ.get("VPS_SSH_USER", "jimmyg")
 VPS_SSH_KEY_PATH = os.environ.get("VPS_SSH_KEY_PATH", "/app/ssh/vps_id_ed25519")
+
+# Telegram notifications
+TELEGRAM_BOT_TOKEN = os.environ.get("TELEGRAM_BOT_TOKEN", "")
+TELEGRAM_CHAT_ID = os.environ.get("TELEGRAM_CHAT_ID", "")
diff --git a/dashboard/backend/main.py b/dashboard/backend/main.py
index ee892bb..52bea23 100644
--- a/dashboard/backend/main.py
+++ b/dashboard/backend/main.py
@@ -3,8 +3,45 @@ from fastapi.staticfiles import StaticFiles
 from routers import docker_router, gitea, files, terminal, system, auth
 import auth as auth_module
 
+import asyncio
+import httpx
+from config import TELEGRAM_BOT_TOKEN, TELEGRAM_CHAT_ID
+
 app = FastAPI(title="NAS Dashboard")
 
+async def monitor_containers():
+    import docker
+    from config import DOCKER_HOST
+    client = docker.DockerClient(base_url=DOCKER_HOST)
+    
+    # Store initial states
+    previous_states = {}
+    
+    while True:
+        try:
+            for c in client.containers.list(all=True):
+                current_status = c.status
+                if c.id in previous_states:
+                    prev_status = previous_states[c.id]
+                    # Alert if it transitioned from running to exited or unhealthy
+                    if prev_status == "running" and current_status in ("exited", "dead"):
+                        msg = f"⚠️ *Container Alert* ⚠️\nContainer `{c.name}` just stopped unexpectedly (Status: {current_status})."
+                        # Use httpx directly as system.send_notification expects a FastAPI request object usually
+                        if TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID:
+                            async with httpx.AsyncClient() as hc:
+                                await hc.post(
+                                    f"https://api.telegram.org/bot{TELEGRAM_BOT_TOKEN}/sendMessage",
+                                    data={"chat_id": TELEGRAM_CHAT_ID, "text": msg, "parse_mode": "Markdown"},
+                                )
+                previous_states[c.id] = current_status
+        except Exception as e:
+            print(f"Error in container monitor: {e}")
+        
+        await asyncio.sleep(60)
+
+@app.on_event("startup")
+async def startup_event():
+    asyncio.create_task(monitor_containers())
 # Public auth endpoints
 app.include_router(auth.router, prefix="/api/auth", tags=["auth"])
 
diff --git a/dashboard/backend/routers/system.py b/dashboard/backend/routers/system.py
index ec193af..0539b5d 100644
--- a/dashboard/backend/routers/system.py
+++ b/dashboard/backend/routers/system.py
@@ -5,12 +5,10 @@ import platform
 import time
 import httpx
 from fastapi import APIRouter
+from config import TELEGRAM_BOT_TOKEN, TELEGRAM_CHAT_ID
 
 router = APIRouter()
 
-TELEGRAM_BOT_TOKEN = os.environ.get("TELEGRAM_BOT_TOKEN", "")
-TELEGRAM_CHAT_ID = os.environ.get("TELEGRAM_CHAT_ID", "")
-
 
 @router.post("/notify")
 async def send_notification(body: dict):