Scrutiny — Smart Hard Drive Monitoring with S.M.A.R.T. Data

Introduction

Scrutiny is a self-hosted hard drive monitoring tool that presents S.M.A.R.T. data in a clean web dashboard. It goes beyond raw attribute numbers by applying vendor-specific failure thresholds and historical trend analysis, helping homelab and server operators catch failing drives before they lose data.

What Scrutiny Does

• Collects S.M.A.R.T. attributes from HDDs, SSDs, and NVMe drives on a schedule
• Displays drive health status with pass, warn, and fail indicators per attribute
• Tracks historical trends for critical metrics like reallocated sectors and temperature
• Sends notifications via email, Discord, Slack, Gotify, and 30+ other services
• Supports multiple hosts by running lightweight collectors that report to a central hub

Architecture Overview

Scrutiny has two components: a collector agent and a web server. The collector runs smartctl on each host and posts JSON results to the web API. The web server is a Go binary that stores metrics in an InfluxDB time-series database and serves the Angular dashboard. The omnibus Docker image bundles both components and InfluxDB for single-host setups.

Self-Hosting & Configuration

• Deploy the omnibus Docker image for single-host setups with all components bundled
• For multi-host monitoring, run the collector image on each machine and point it to the central hub
• Map physical drives into the container with --device flags and add SYS_RAWIO capability
• Configure notification endpoints in scrutiny.yaml using the Shoutrrr URL format
• Adjust scan schedules via cron expressions in the configuration file

Key Features

• Vendor-specific S.M.A.R.T. threshold analysis for more accurate failure prediction
• Historical charts showing attribute trends over weeks and months
• Hub-and-spoke architecture for monitoring drives across multiple servers
• Notification support for 30+ services via the Shoutrrr library
• NVMe support alongside traditional SATA and SAS drives

Comparison with Similar Tools

• Netdata — general infrastructure monitoring; Scrutiny specializes in drive health with deeper S.M.A.R.T. analysis
• smartmontools alone — CLI-only with no dashboard or history; Scrutiny adds visualization and alerting
• Disk Utility (GNOME) — shows current S.M.A.R.T. status but has no trend tracking or remote monitoring
• Hard Disk Sentinel — commercial Windows tool; Scrutiny is free and runs on Linux servers

FAQ

Q: What drives does Scrutiny support? A: SATA HDDs, SATA SSDs, SAS drives, and NVMe devices. Any drive supported by smartctl should work.

Q: Can I monitor drives on multiple servers? A: Yes. Run the lightweight collector container on each host and configure it to report to a central Scrutiny web instance.

Q: How often are drives scanned? A: By default every 24 hours. You can adjust the cron schedule in the configuration to scan more or less frequently.

Q: Does Scrutiny require InfluxDB? A: The omnibus image bundles InfluxDB automatically. For split deployments, you point the web component at an external InfluxDB instance.

Sources

• https://github.com/AnalogJ/scrutiny
• https://github.com/AnalogJ/scrutiny/blob/master/docs/INSTALL_DOCKER.md