fix(alerts): auto-resolve device_down alerts when agent polls device successfully

When agents poll devices via SNMP and get successful results, the code was
only processing sensor/interface data without updating device status or
resolving alerts. This caused device_down alerts to persist even after
devices came back online.

This fix:
- Updates device status to 'up' when SNMP polling succeeds
- Auto-resolves any active device_down alerts
- Creates device_up alert (matching DeviceMonitorWorker behavior)
- Broadcasts device_status_changed via PubSub

Also fixes device_up alert creation to include resolved_at timestamp
(matching DeviceMonitorWorker behavior).

Test coverage: Added test verifying auto-resolution when device recovers.

lib/towerops_web/channels/agent_channel.ex

@@ -1468,6 +1468,7 @@ defmodule ToweropsWeb.AgentChannel do
       device_id: device.id,
       alert_type: :device_up,
       triggered_at: now,
+      resolved_at: now,
       message: message
     })
 
@@ -1604,6 +1605,21 @@ defmodule ToweropsWeb.AgentChannel do
   defp process_polling_result(device, result, _socket) do
     Devices.update_snmp_poll_time(device)
 
+    # Successful SNMP poll means device is up - check for status change and resolve alerts
+    old_status = device.status
+    new_status = :up
+
+    if old_status != new_status do
+      Devices.update_device_status(device, new_status)
+      handle_agent_status_change(device, old_status, new_status)
+
+      Phoenix.PubSub.broadcast(
+        Towerops.PubSub,
+        "device:#{device.id}",
+        {:device_status_changed, device.id, new_status, nil}
+      )
+    end
+
     snmp_device = device.snmp_device
     oid_values = normalize_oid_keys(result.oid_values)
     # Use server time to avoid clock drift issues between agents

test/towerops_web/channels/agent_channel_test.exs

@@ -916,6 +916,60 @@ defmodule ToweropsWeb.AgentChannelTest do
       updated_sensor = Towerops.Repo.get!(Sensor, sensor_with_divisor.id)
       assert updated_sensor.last_value == 12.0
     end
+
+    test "poll result auto-resolves device_down alert when device comes back up", %{
+      socket: socket,
+      discovered_device: device,
+      sensor: sensor
+    } do
+      # Set device status to down
+      {:ok, device} = Towerops.Devices.update_device_status(device, :down)
+
+      # Create a device_down alert
+      {:ok, alert} =
+        Towerops.Alerts.create_alert(%{
+          device_id: device.id,
+          alert_type: "device_down",
+          triggered_at: DateTime.utc_now(),
+          message: "Device is not responding to SNMP"
+        })
+
+      assert is_nil(alert.resolved_at)
+      assert device.status == :down
+
+      # Send successful SNMP polling result
+      result = %SnmpResult{
+        device_id: device.id,
+        job_type: :POLL,
+        job_id: "poll:#{device.id}",
+        timestamp: DateTime.to_unix(DateTime.utc_now()),
+        oid_values: %{
+          sensor.sensor_oid => "45"
+        }
+      }
+
+      push(socket, "result", encode_payload(result))
+      Process.sleep(50)
+
+      # Verify device status updated to up
+      updated_device = Towerops.Devices.get_device(device.id)
+      assert updated_device.status == :up
+
+      # Verify device_down alert was auto-resolved
+      updated_alert = Towerops.Alerts.get_alert(alert.id)
+      assert updated_alert.resolved_at
+
+      # Verify device_up alert was created
+      up_alerts =
+        device.id
+        |> Towerops.Alerts.list_devices_alerts()
+        |> Enum.filter(&(&1.alert_type == "device_up"))
+
+      assert length(up_alerts) == 1
+      up_alert = hd(up_alerts)
+      assert up_alert.message == "Device is now responding to SNMP"
+      assert up_alert.resolved_at
+    end
   end
 
   # ── Stage 10: SNMPv3 credential path ────────────────────────────────