Add host-poller: chassis IPMI/iDRAC + CPU + on-metal drive temps

New third producer container (host-poller) owning the server chassis
subsystem, separate from the SAS-shelf poller:
- services/host_sensors.py: in-band IPMI (ipmitool sdr) for Inlet/Exhaust/etc.
  + CPU package temps from coretemp hwmon
- host_poller.py: own single-instance lock + heartbeat (host_poll keys),
  restart-storm guard, paced; writes jbod:host_sensors + jbod:host_drives
- move host/on-metal drive collection off the SAS poller to host-poller
  (reads jbod:zfs_map for annotation)
- store: generalized lock/meta with a key param; host_sensors + host-poll keys
- mqtt_publisher: publish env/CPU/host-drive temps as entities on the hub
  device (JBOD Monitor)
- Dockerfile host-poller target (ipmitool+smartmontools); compose.infra adds
  host-poller; build.sh gains host-poller/all

services/mqtt_publisher.py

@@ -19,6 +19,7 @@ import logging
 import os
 import re
 
+from services import store
 from services.secrets import read_kv2
 from services.temps import build_temp_snapshot, get_node_id
 
@@ -242,6 +243,61 @@ class MqttPublisher:
             self._publish_discovery(enc)
             self._publish_state(enc)
 
+    def publish_host(self, sensors: dict, drives: list) -> None:
+        """Publish chassis temps (IPMI env, CPU, on-metal drives) on the hub."""
+        self._publish_hub()
+        state_topic = f"{self.base_topic}/{self.node}/host/state"
+        device = self._hub_device_block()
+        payload: dict = {}
+
+        def announce(object_id: str, cfg: dict) -> None:
+            uid = cfg["unique_id"]
+            if uid not in self._discovered:
+                topic = f"{self.discovery_prefix}/sensor/{self.hub_id}/{object_id}/config"
+                self._client.publish(topic, json.dumps(cfg), qos=1, retain=True)
+                self._discovered.add(uid)
+
+        common = {
+            "device_class": "temperature",
+            "unit_of_measurement": "°C",
+            "state_class": "measurement",
+            "state_topic": state_topic,
+            "availability_topic": self.availability_topic,
+            "device": device,
+        }
+
+        def add(kind: str, name: str, temp, icon: str | None = None) -> None:
+            if temp is None or not name:
+                return
+            key = f"{kind}_{_slug(name)}"
+            payload[key] = temp
+            cfg = {
+                **common,
+                "name": name,
+                "unique_id": f"{self.hub_id}_{key}",
+                "value_template": f"{{{{ value_json.{key} }}}}",
+            }
+            if icon:
+                cfg["icon"] = icon
+            announce(key, cfg)
+
+        for s in sensors.get("env", []):
+            add("env", s.get("name"), s.get("temperature_c"))
+        for c in sensors.get("cpu", []):
+            add("cpu", c.get("name"), c.get("temperature_c"), icon="mdi:cpu-64-bit")
+
+        def add_drive(d: dict) -> None:
+            dev = d.get("device")
+            if dev:
+                add("drive", f"Drive {dev}", d.get("temperature_c"), icon="mdi:harddisk")
+
+        for d in drives:
+            add_drive(d)
+            for pd in d.get("physical_drives", []):
+                add_drive(pd)
+
+        self._client.publish(state_topic, json.dumps(payload), qos=0, retain=True)
+
     async def run(self) -> None:
         """Background loop: build a snapshot and publish on an interval."""
         await asyncio.sleep(3)  # let the first SMART poll populate the cache
@@ -249,9 +305,14 @@ class MqttPublisher:
             try:
                 snapshot = await build_temp_snapshot()
                 self.publish_snapshot(snapshot)
+                host_sensors = await store.get_host_sensors() or {}
+                host_drives = await store.get_host_drives() or []
+                self.publish_host(host_sensors, host_drives)
                 logger.info(
-                    "MQTT published temps for %d enclosure(s)",
+                    "MQTT published temps for %d enclosure(s) + host (%d env, %d cpu)",
                     len(snapshot.get("enclosures", [])),
+                    len(host_sensors.get("env", [])),
+                    len(host_sensors.get("cpu", [])),
                 )
             except asyncio.CancelledError:
                 raise