19552 Redfish: add new checks and bugfixes from PR #892

Add PDU status, power consumption, power redundancy, and storage
battery monitoring. Enhance ethernet interfaces with link change
detection, add configurable drive/volume naming schemes, and fix
SSD endurance reporting. Harden special agents against missing
Redfish endpoints.

Closes: #892
Jira-Ref: CMK-32388
Change-Id: Id0122e82b7eeef1ddfb099584269e8f372e30624
Co-authored-by: Sebastian Groemcke <sebastian.groemcke@checkmk.com>

Author: Yogibaer75

.werks/19552.md

@@ -0,0 +1,40 @@
+[//]: # (werk v3)
+# Redfish: add PDU, power consumption, and power redundancy monitoring
+
+key        | value
+---------- | ---
+date       | 2026-03-24T08:57:36.083866+00:00
+version    | 2.5.0b2
+class      | feature
+edition    | community
+component  | checks
+level      | 2
+compatible | yes
+
+The Redfish integration now supports several new monitoring capabilities and improvements.
+
+## New checks
+
+- _PDU %s_: Monitors Power Distribution Unit status including firmware, serial number, model, and health state.
+- _Power consumption %s_: Tracks average, minimum, and maximum power consumption with per-system metrics and graphs.
+- _Power redundancy %s_: Verifies power supply redundancy configuration and health.
+- _Storage controller %s battery_: Monitors Dell storage controller battery status (Dell OEM).
+
+## Enhanced checks
+
+- _Physical port %s_: Now detects link status and speed changes compared to what was seen at discovery. Alert severity is configurable via the new _Redfish Ethernet Interface_ ruleset (default: `CRIT` for link status change, `WARN` for speed change). A (periodic) service rediscovery resets the baseline values.
+- _Storage controller %s_: New _Redfish Storage Controller_ ruleset lets you choose between full detail and rollup-only health reporting.
+- _Drive %s_ and _Volume %s_: New discovery rulesets let you choose between classic item naming and a structured controller-ID format.
+- _System state %s_: Now shows serial number, SKU, and Dell OEM rollup details.
+
+## Bugfixes
+
+- _Drive %s_ (physical drives): SSD endurance now correctly displays "Media Life Left" instead of raw utilization percentage.
+- The Redfish power agent now fetches chassis sensor data (temperature, humidity) from PDUs. Previously, only outlets and mains were collected.
+- The Redfish special agents no longer crash when a system, chassis, or power equipment endpoint is missing or a section has no data.
+
+## Inventory
+
+A new Redfish drives inventory view is available under _Hardware > Storage > Redfish drives_, showing manufacturer, model, serial number, firmware version, capacity, and media type.
+
+Thanks to Andreas Döhler for the contribution.

packages/cmk-plugins/BUILD

@@ -386,6 +386,7 @@ py_library(
         ":lib",
         "//packages/cmk-plugin-apis:agent_based",
         "//packages/cmk-plugin-apis:graphing",
+        "//packages/cmk-plugin-apis:inventory_ui",
         "//packages/cmk-plugin-apis:password_store",
         "//packages/cmk-plugin-apis:rulesets",
         "//packages/cmk-plugin-apis:server_side_calls",

packages/cmk-plugins/cmk/plugins/redfish/agent_based/inv_redfish_firmware_hpe_ilo4.py

@@ -0,0 +1,63 @@
+#!/usr/bin/env python3
+# Copyright (C) 2026 Checkmk GmbH - License: GNU General Public License v2
+# This file is part of Checkmk (https://checkmk.com). It is subject to the terms and
+# conditions defined in the file COPYING, which is part of this source code package.
+"""Redfish storage drives inventory plugin.
+
+Extracts hardware component data from Redfish drive sections to build
+a consolidated inventory view of physical drives.
+"""
+
+from cmk.agent_based.v2 import (
+    InventoryPlugin,
+    InventoryResult,
+    TableRow,
+)
+from cmk.plugins.redfish.lib import RedfishAPIData
+
+
+def _extract_component_name(entry: RedfishAPIData) -> str:
+    """Build a component name from entry data."""
+    entry_id = str(entry.get("Id", ""))
+    name = str(entry.get("Name", ""))
+    if entry_id and name and entry_id != name:
+        return f"{entry_id}-{name}"
+    return name or entry_id or "Unknown"
+
+
+def inventorize_redfish_drives(section: RedfishAPIData) -> InventoryResult:
+    """Create inventory entries from Redfish drive data."""
+    for _key, entry in section.items():
+        if not isinstance(entry, dict):
+            continue
+        component_name = _extract_component_name(entry)
+
+        inventory_columns: dict[str, int | float | str | bool | None] = {}
+        if manufacturer := entry.get("Manufacturer"):
+            inventory_columns["manufacturer"] = manufacturer
+        if model := entry.get("Model"):
+            inventory_columns["model"] = model
+        if serial := entry.get("SerialNumber"):
+            inventory_columns["serial"] = serial
+        if firmware := entry.get("FirmwareVersion"):
+            inventory_columns["firmware_version"] = firmware
+        if capacity := entry.get("CapacityBytes"):
+            inventory_columns["capacity_bytes"] = capacity
+        if media_type := entry.get("MediaType"):
+            inventory_columns["media_type"] = media_type
+
+        if not inventory_columns:
+            continue
+
+        yield TableRow(
+            path=["hardware", "storage", "redfish_drives"],
+            key_columns={"component": component_name},
+            inventory_columns=inventory_columns,
+        )
+
+
+inventory_plugin_redfish_storage_inventory = InventoryPlugin(
+    name="redfish_storage_inventory",
+    sections=["redfish_drives"],
+    inventory_function=inventorize_redfish_drives,
+)

packages/cmk-plugins/cmk/plugins/redfish/agent_based/inventory_redfish_data.py

@@ -3,6 +3,9 @@
 # This file is part of Checkmk (https://checkmk.com). It is subject to the terms and
 # conditions defined in the file COPYING, which is part of this source code package.
 
+from collections.abc import Mapping
+from typing import Any
+
 from cmk.agent_based.v2 import (
     AgentSection,
     CheckPlugin,
@@ -26,21 +29,53 @@
 )
 
 
-def discovery_redfish_drives(section: RedfishAPIData) -> DiscoveryResult:
-    for key in section.keys():
+def _build_drive_item(data: RedfishAPIData) -> tuple[str, str]:
+    """Build classic and ctrlid item names for a drive.
+
+    Classic: "Id-Name" (e.g., "0-1.2TB 12G SAS HDD")
+    Ctrlid: structured from OData path (e.g., "0:1:4" from
+            /redfish/v1/Systems/0/Storage/1/Drives/4)
+    """
+    classic = data.get("Id", "0") + "-" + data.get("Name", "")
+    odata_id = data.get("@odata.id", "")
+    if odata_id:
+        parts = odata_id.strip("/").split("/")
+        # Try to extract system:storage:drive or chassis:drive format
+        if len(parts) >= 7 and "Drives" in parts:
+            drive_idx = parts.index("Drives")
+            if drive_idx >= 4:
+                # e.g., /redfish/v1/Systems/0/Storage/1/Drives/4 -> 0:1:4
+                ctrlid = ":".join(parts[3:drive_idx:2] + [parts[drive_idx + 1]])
+            else:
+                ctrlid = ":".join(parts[-2:])
+        elif len(parts) >= 5:
+            # e.g., /redfish/v1/Chassis/DE00B000/Drives/0 -> DE00B000:0
+            ctrlid = ":".join(parts[-2:])
+        else:
+            ctrlid = classic
+    else:
+        ctrlid = classic
+    return classic, ctrlid
+
+
+def discovery_redfish_drives(params: Mapping[str, Any], section: RedfishAPIData) -> DiscoveryResult:
+    naming = params.get("item", "classic")
+    for key in section:
         if section[key].get("Status", {}).get("State") == "Absent":
             continue
-        if not section[key]["Name"]:
+        if not section[key].get("Name"):
             continue
-        item = section[key].get("Id", "0") + "-" + section[key]["Name"]
+        classic, ctrlid = _build_drive_item(section[key])
+        item = ctrlid if naming == "ctrlid" else classic
         yield Service(item=item)
 
 
 def check_redfish_drives(item: str, section: RedfishAPIData) -> CheckResult:
     data = None
-    for key in section.keys():
-        if item == section[key].get("Id", "0") + "-" + section[key]["Name"]:
-            data = section.get(key, None)
+    for key in section:
+        classic, ctrlid = _build_drive_item(section[key])
+        if item in (classic, ctrlid):
+            data = section.get(key)
             break
     if data is None:
         return
@@ -70,5 +105,7 @@ def check_redfish_drives(item: str, section: RedfishAPIData) -> CheckResult:
     service_name="Drive %s",
     sections=["redfish_drives"],
     discovery_function=discovery_redfish_drives,
+    discovery_ruleset_name="discovery_redfish_drives",
+    discovery_default_parameters={"item": "classic"},
     check_function=check_redfish_drives,
 )

packages/cmk-plugins/cmk/plugins/redfish/agent_based/redfish_drives.py

@@ -3,10 +3,6 @@
 # This file is part of Checkmk (https://checkmk.com). It is subject to the terms and
 # conditions defined in the file COPYING, which is part of this source code package.
 
-# mypy: disable-error-code="comparison-overlap"
-
-# mypy: disable-error-code="unreachable"
-
 from collections.abc import Mapping
 from typing import Any
 
@@ -32,12 +28,19 @@
 )
 
 
+def _render_speed(speed: int) -> str:
+    """Render link speed with appropriate unit."""
+    if speed >= 1000:
+        return f"{speed / 1000:g} Gbps"
+    return f"{speed} Mbps"
+
+
 def discovery_redfish_ethernetinterfaces(
     params: Mapping[str, Any], section: RedfishAPIData
 ) -> DiscoveryResult:
     """Discover single interfaces"""
     disc_state = params.get("state")
-    for key in section.keys():
+    for key in section:
         if not section[key].get("Status"):
             continue
         if section[key].get("Status", {}).get("State") in [
@@ -52,47 +55,66 @@ def discovery_redfish_ethernetinterfaces(
             continue
         if section[key].get("LinkStatus", "NOLINK") in ["LinkUp"] and disc_state == "down":
             continue
-        yield Service(item=section[key]["Id"])
 
+        speed = section[key].get("SpeedMbps", 0)
+        if speed == 0:
+            speed = section[key].get("CurrentLinkSpeedMbps", 0)
+
+        yield Service(
+            item=section[key]["Id"],
+            parameters={
+                "discover_speed": speed if speed else 0,
+                "discover_link_status": section[key].get("LinkStatus", "NOLINK"),
+            },
+        )
 
-def check_redfish_ethernetinterfaces(item: str, section: RedfishAPIData) -> CheckResult:
+
+def check_redfish_ethernetinterfaces(
+    item: str, params: Mapping[str, Any], section: RedfishAPIData
+) -> CheckResult:
     """Check single interfaces"""
-    data = section.get(item, None)
+    data = section.get(item)
     if data is None:
         return
 
+    # Link status
+    link_state = State.OK
+    link_summary = "Link: No info"
+    if (link_status := data.get("LinkStatus")) is not None:
+        link_summary = f"Link: {link_status}"
+        discover_link = params.get("discover_link_status")
+        if discover_link and discover_link != link_status:
+            link_state = State(params.get("state_if_link_status_changed", 2))
+            link_summary = f"Link: {link_status} (changed from {discover_link})"
+    yield Result(state=link_state, summary=link_summary)
+
+    # Speed
+    link_speed = data.get("CurrentLinkSpeedMbps") or data.get("SpeedMbps") or 0
+    speed_state = State.OK
+    speed_summary = f"Speed: {_render_speed(link_speed)}" if link_speed else "Speed: Unknown"
+    discover_speed = params.get("discover_speed")
+    if discover_speed and link_speed and discover_speed != link_speed:
+        speed_state = State(params.get("state_if_link_speed_changed", 1))
+        speed_summary = (
+            f"Speed: {_render_speed(link_speed)} (changed from {_render_speed(discover_speed)})"
+        )
+    yield Result(state=speed_state, summary=speed_summary)
+
+    # MAC address
     mac_addr = ""
     if data.get("AssociatedNetworkAddresses"):
         mac_addr = ", ".join(data.get("AssociatedNetworkAddresses"))
     elif data.get("MACAddress"):
         mac_addr = data.get("MACAddress")
+    if mac_addr:
+        yield Result(state=State.OK, summary=f"MAC: {mac_addr}")
 
-    link_speed = 0
-    if data.get("CurrentLinkSpeedMbps"):
-        link_speed = data.get("CurrentLinkSpeedMbps")
-    elif data.get("SpeedMbps"):
-        link_speed = data.get("SpeedMbps")
-    if link_speed is None:
-        link_speed = 0
-
-    link_status = "Unknown"
-    if data.get("LinkStatus"):
-        link_status = data.get("LinkStatus")
-        if link_status is None:
-            link_status = "Down"
-
-    int_msg = f"Link: {link_status}, Speed: {link_speed:0.0f}Mbps, MAC: {mac_addr}"
-    yield Result(state=State(0), summary=int_msg)
-
+    # Health state
     if data.get("Status"):
         dev_state, dev_msg = redfish_health_state(data.get("Status", {}))
-        status = dev_state
-        message = dev_msg
+        yield Result(state=State(dev_state), notice=dev_msg)
     else:
-        status = 0
-        message = "No known status value found"
-
-    yield Result(state=State(status), notice=message)
+        yield Result(state=State.OK, notice="No known status value found")
 
 
 check_plugin_redfish_ethernetinterfaces = CheckPlugin(
@@ -103,4 +125,6 @@ def check_redfish_ethernetinterfaces(item: str, section: RedfishAPIData) -> Chec
     discovery_ruleset_name="discovery_redfish_ethernetinterfaces",
     discovery_default_parameters={"state": "updown"},
     check_function=check_redfish_ethernetinterfaces,
+    check_ruleset_name="check_redfish_ethernetinterfaces",
+    check_default_parameters={},
 )