host inventory

Pre-Alpha. This page describes behavior that may change.

The host inventory plugin provides hardware detection, SMART health monitoring, and a hardware tuning engine. It exposes system hardware state through CLI commands and Prometheus metrics.

CLI

show host                  # Full hardware inventory (JSON)
show host cpu              # CPU details (model, cores, frequency)
show host nic              # Network interface hardware (speed, carrier, firmware)
show host memory           # DIMM layout, ECC status
show host dmi              # DMI/SMBIOS data (vendor, model, serial)
show host thermal          # Temperature sensors
show host storage          # Disk capacity, SMART health, wear level
show host kernel           # Kernel version, loaded modules
show host all              # All sections combined
show host --text [section] # Human-readable summary

Prometheus metrics

Metrics are exposed under the ze_host_* namespace when the telemetry prometheus block is configured:

Metric	Description
`ze_host_memory_*`	DIMM size, ECC errors, type.
`ze_host_cpu_*`	Core count, frequency, model.
`ze_host_nic_*`	Speed, carrier, firmware version.
`ze_host_thermal_*`	Temperature readings per sensor.
`ze_host_storage_*`	SMART health status, capacity, wear level.

Hardware tuning

The tuning engine applies system optimizations at startup and on config reload. It adjusts kernel parameters (sysctl), interrupt affinity, and NIC ring buffer sizes based on the detected hardware profile.