feat(monitoring): add update-run log, fix scheduler skipping stale customer containers

The automatic scheduler only recreated customer containers when it had just
found a NEW image on Docker Hub. If an image was already pulled locally
(e.g. from an earlier manual pull, or auto-apply was off at the time) but
customer containers hadn't been recreated yet, the daily check reported
"already up to date" and silently never touched those customers again —
the recreate step needs to run independently of whether Hub had anything
new, based on each customer's actual container-vs-local-image state.

Also adds a persistent UpdateRunLog: every scheduled and manual pull/update
run now records per-image pull results and per-customer recreate results,
visible in Monitoring > Update Log, instead of only existing in container
logs.

app/routers/deployments.py

@@ -9,7 +9,7 @@ from sqlalchemy.orm import Session
 from app.database import SessionLocal, get_db
 from app.dependencies import get_current_user
 from app.models import Customer, Deployment, SystemConfig, User
-from app.services import docker_service, image_service, netbird_client_update_service, netbird_service
+from app.services import docker_service, image_service, netbird_client_update_service, netbird_service, update_log_service
 from app.utils.security import decrypt_value, encrypt_value
 from app.utils.validators import NetbirdApiTokenPayload, NetbirdClientAutoUpdatePayload
 
@@ -246,9 +246,18 @@ async def update_customer_images(
             status_code=status.HTTP_503_SERVICE_UNAVAILABLE, detail="System not configured."
         )
 
+    started_at = datetime.utcnow()
     instance_dir = f"{config.data_dir}/{customer.subdomain}"
     result = await image_service.update_customer_containers(instance_dir, deployment.container_prefix)
 
+    update_log_service.record_run(
+        db, run_type="manual", trigger=current_user.username, started_at=started_at,
+        apply_attempted=True,
+        customer_results=[{
+            "customer_name": customer.name, "success": result["success"], "error": result.get("error"),
+        }],
+    )
+
     if not result["success"]:
         raise HTTPException(
             status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,