SMARTctl Exporter Dashboard

Dashboard for prometheus-community/smartctl_exporter, which wraps smartctl output for each block device into Prometheus metrics via SMART/NVMe self-monitoring data.

Panels

• Disk Temperature — per-device temperature over time, thresholded at 80°C
• Devices / Device Count by Interface (pie) — inventory of monitored devices, split by interface (sat, nvme, scsi, etc.)
• Media Errors / Critical Warnings — NVMe media error and critical warning counters, only plotted when non-zero
• Unhealthy — count of devices where smartctl_device_smart_status reports failed
• Disk Lifetime — table of power-on hours and power-cycle count per device
• Total Data Written on SSDs — cumulative bytes written, computed from smartctl_device_bytes_written where available, falling back to Total_LBAs_Written × logical block size for devices that don't expose bytes written directly
• Wear Leveling Over Time — SSD wear level as a percentage over time: 100 − Wear_Leveling_Count for SATA drives, smartctl_device_percentage_used for NVMe

Variables

Template variables (node, disk, type, device_model, serial_number) filter every panel down to a specific host, device, interface, model, or serial — all multi-select with "All" supported.

Data collection

Requires smartctl_exporter running with access to the target block devices (typically via a privileged container/host process, since smartctl needs raw device access) and a Prometheus scrape job pointed at it:

```yaml scrape_configs:

• job_name: smartctl-exporter static_configs:
  • targets: ['smartctl-exporter:9633'] ```

No CollectD involved — this exporter talks directly to smartctl and exposes /metrics for Prometheus scraping.