19472: Fix false CRIT for NetApp temperature sensors without transceivers

DavidGerva · DavidGerva · commit 52bc95cc9b94 · 2026-04-08T11:30:39.000+02:00
Only monitor sensors that report an actual temperature value and
at least one threshold.

SUP-28183

Change-Id: I9b68380a56a966350135a80950b6a440ad48986a
diff --git a/.werks/19472.md b/.werks/19472.md
@@ -0,0 +1,29 @@
+[//]: # (werk v3)
+# Fix false CRIT for NetApp temperature sensors without transceivers
+
+key        | value
+---------- | ---
+date       | 2026-04-01T16:17:31.156156+00:00
+version    | 2.5.0b4
+class      | fix
+edition    | community
+component  | checks
+level      | 1
+compatible | yes
+
+The `netapp_ontap_temp` check incorrectly reported CRIT for temperature sensors
+on NetApp ONTAP systems where DAC cables are used
+instead of optical SFP transceivers, or where SSDs have no temperature sensors
+installed.
+
+NetApp firmware reports these ports as `installed: true` even when no transceiver
+is present, and returns `state: error` with `temperature: null` and all thresholds
+`null`. This seems a firmware/API limitation: the hardware detection does not
+distinguish between "not installed" and "error".
+
+The check now only monitors sensors that report an actual temperature value and
+at least one threshold. DAC-only sensors are silently ignored, eliminating the
+false CRIT state.
+
+If you have existing services showing CRIT for such sensors, they will disappear
+after the next check.
diff --git a/packages/cmk-plugins/cmk/plugins/netapp/models.py b/packages/cmk-plugins/cmk/plugins/netapp/models.py
@@ -542,6 +542,27 @@ class ShelfTemperatureModel(ShelfObjectModel):
     high_warning: int | None = None
     high_critical: int | None = None
 
+    def _has_sensor_data(self) -> bool:
+        """Return True only if the sensor reports a real temperature AND at least one threshold.
+
+        NetApp firmware reports installed=true for DAC-only ports (no SFP transceiver present),
+        setting state=error and all values to null. These are not real errors and must be skipped.
+        This appears as a firmware/API bug in NetApp: the hardware detection does not differentiate
+        between "not installed" and "error".
+        """
+        if self.temperature is None:
+            return False
+
+        return None not in (
+            self.low_warning,
+            self.low_critical,
+            self.high_warning,
+            self.high_critical,
+        )
+
+    def consider_installed(self) -> bool:
+        return super().consider_installed() and self._has_sensor_data()
+
 
 class ShelfPsuModel(ShelfObjectModel):
     """
diff --git a/packages/cmk-plugins/tests/cmk/plugins/netapp/agent_based/test_netapp_ontap_temp.py b/packages/cmk-plugins/tests/cmk/plugins/netapp/agent_based/test_netapp_ontap_temp.py
@@ -18,6 +18,7 @@
 from cmk.plugins.netapp.agent_based.netapp_ontap_temp import (
     _check_netapp_ontap_temp,
     discovery_netapp_ontap_temp,
+    parse_netapp_ontap_temp,
 )
 from cmk.plugins.netapp.models import ShelfTemperatureModel
 
@@ -32,79 +33,68 @@ class ShelfTemperatureModelFactory(ModelFactory):
             list_id="2",
             state="ok",
             id=111,
-            temperature=20.0,
+            temperature=20,
             ambient=True,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=55,
+            high_critical=60,
         ),
     ],
     "Internal Shelf 2": [
         ShelfTemperatureModelFactory.build(
             list_id="2",
             state="ok",
-            temperature=50.0,
+            temperature=50,
             ambient=False,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=95,
+            high_critical=105,
         ),
     ],
     "Ambient Shelf 1": [
         ShelfTemperatureModelFactory.build(
             list_id="1",
             state="ok",
-            temperature=50.0,
+            temperature=50,
             ambient=True,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=95,
+            high_critical=105,
         ),
         ShelfTemperatureModelFactory.build(
             list_id="1",
             state="ok",
-            temperature=10.0,
+            temperature=10,
             ambient=True,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=95,
+            high_critical=105,
         ),
         ShelfTemperatureModelFactory.build(
             list_id="1",
             state="ok",
-            temperature=20.0,
+            temperature=20,
             ambient=True,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
-        ),
-        ShelfTemperatureModelFactory.build(
-            list_id="1",
-            id=20,
-            state="error",
-            temperature=None,
-            ambient=True,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=95,
+            high_critical=105,
         ),
     ],
     "Internal Shelf 1": [
         ShelfTemperatureModelFactory.build(
             list_id="1",
             state="ok",
-            temperature=30.0,
+            temperature=30,
             ambient=False,
-            low_warning=None,
-            low_critical=None,
-            high_warning=None,
-            high_critical=None,
+            low_warning=5,
+            low_critical=0,
+            high_warning=95,
+            high_critical=105,
         ),
     ],
 }
@@ -142,7 +132,6 @@ def test_check_netapp_ontap_temp_() -> None:
         Metric("temp", 50.0),
         Result(state=State.OK, summary="Average: 26.7 °C"),
         Result(state=State.OK, summary="Lowest: 10 °C"),
-        Result(state=State.CRIT, summary="Additional failed sensors: 1 (1/20)"),
     ]
 
 
@@ -218,3 +207,35 @@ def test_check_netapp_ontap_temp_trend(
         )
 
         assert result[-1] == expected_trend_result
+
+
+@pytest.mark.parametrize(
+    "json_line, expected_count",
+    [
+        pytest.param(
+            # Valid sensor: temperature and thresholds present
+            '{"list_id":"1","id":8,"state":"ok","installed":true,"temperature":57,"ambient":false,'
+            '"low_warning":5,"low_critical":0,"high_warning":95,"high_critical":105}',
+            1,
+            id="valid sensor is kept",
+        ),
+        pytest.param(
+            # DAC-only sensor: installed=true but no SFP — all values null (NetApp firmware bug)
+            '{"list_id":"1","id":10,"state":"error","installed":true,"temperature":null,"ambient":false,'
+            '"low_warning":null,"low_critical":null,"high_warning":null,"high_critical":null}',
+            0,
+            id="DAC sensor with null temperature and null thresholds is filtered",
+        ),
+        pytest.param(
+            # Sensor with temperature but no thresholds is filtered
+            '{"list_id":"1","id":12,"state":"ok","installed":true,"temperature":45,"ambient":false,'
+            '"low_warning":null,"low_critical":null,"high_warning":null,"high_critical":null}',
+            0,
+            id="sensor with temperature but no thresholds is filtered",
+        ),
+    ],
+)
+def test_parse_netapp_ontap_temp_filters_dac_sensors(json_line: str, expected_count: int) -> None:
+    section = parse_netapp_ontap_temp([[json_line]])
+    total = sum(len(sensors) for sensors in section.values())
+    assert total == expected_count
