Scrutiny smart home integration, free and open source
Disk health from Scrutiny (S.M.A.R.T.) in Gladys: state, temperature, and scenes when a disk fails.
Published by guim31
- Devices
- Version 1.0.1
- Requires Gladys >=5.1.0
- ★ 0
Scrutiny watches the health of your hard drives and SSDs: its collector reads their S.M.A.R.T. data (usually once a day) and its web dashboard tells which disk passes, which one shows a warning and which one fails, with thresholds drawn from real failure rates. This integration brings that dashboard into Gladys: each disk becomes a device with its health, temperature and counters, scenes react when a disk fails, passes again or is no longer collected, and a dashboard widget lists the disks that need attention first.
Gladys reads the Scrutiny web server: it never touches your disks. Everything stays on your network (no cloud).
Tested against a real Scrutiny in Docker (v0.9.5, the omnibus image), fed with
real smartctl outputs from the Scrutiny test suite: disks passed, with a
warning, failed by S.M.A.R.T. and by Scrutiny, an NVMe disk, a stale
collection. Not yet on a live installation: feedback welcome on
the issues.
What you need
- Gladys Assistant 5.1.0 or later.
- A Scrutiny web server (the
omnibusimage, orweb+ InfluxDB) reachable from Gladys, usually on port 8080, with at least one collector sending it data. Scrutiny v0.8 and v0.9 are supported. - If a reverse proxy asks for a login in front of Scrutiny: its user and password (basic authentication). Scrutiny itself has no login.
Getting started
1. Find the address of Scrutiny
It is the address of its dashboard, for example http://192.168.1.2:8080. A
URL copied from the browser works too (…/web/dashboard): the integration
keeps the server part. Behind a reverse proxy with a sub-path (the basepath
setting of Scrutiny), enter the address with that sub-path.
2. Configure the integration
In Gladys, open Integrations → Scrutiny → Configuration:
- Scrutiny address: the address found above.
- Proxy user (optional) and Proxy password (optional): only if a reverse proxy asks for them.
- Refresh interval: how often Scrutiny is read, 15 minutes by default. One request for the list of disks and one for the settings each time; the attributes of a disk are read again only after a new collection.
- Collection too old after: after how long without a new collection a disk fires "Disk no longer collected" (36 hours by default, for a collector that runs once a day).
- Language of the names: the language of the feature names (set when a device is added) and of the texts sent to scenes.
Then click Test the connection: it tells how many disks Scrutiny follows, and how many failed.
3. Add the disks
Open the Discover tab of the integration and add the disks to follow, and the Scrutiny device if you want the totals. Archived disks of Scrutiny are not proposed. A disk is identified by its Scrutiny id (built from its model, serial number and WWN: a WWN alone is not always unique): moving it to another port or machine keeps the same device, and so does an upgrade from Scrutiny v0.8 to v0.9.
The devices
Each disk has these features:
- Disk failed: yes when Scrutiny marks the disk as failed, the way its dashboard does (the "status threshold" of the Scrutiny settings: S.M.A.R.T., Scrutiny or both). The condition to use in scenes.
- Health: OK, Warning (the disk passes but an attribute does not look good) or Failed, as a text.
- Temperature: at the last collection, in °C.
- Power-on hours: as reported by the disk.
- Reallocated sectors and Pending sectors: S.M.A.R.T. attributes 5 and 197, only on the disks that report them (SATA). A growing count is the first sign of a dying disk.
- Since last collection: hours since the collector last sent data for the disk.
The Scrutiny device counts the Failed disks, the Disks with a warning and the Disks in total.
Dashboard
The Scrutiny disks widget shows three tiles (failed, warnings, the hottest temperature), then one row per disk: the failed ones first, then the warnings, the disks no longer collected and the healthy ones, ten rows at most. Choose some disks in its settings to show only them. Refresh reads Scrutiny again.
Scenes
Triggers
- A disk fails: Scrutiny marks a disk as failed after a collection. Variables: the disk name, its model, the failing attributes and its temperature.
- A disk passes again: a failed disk passes again at a new collection, or after a change of the Scrutiny status threshold. Variables: the disk name, its model and how many hours it was failed.
- Disk no longer collected: no new collection for a disk for the delay of the configuration (the collector stopped, the disk is gone). One event per disk. Variables: the disk name, its model and the hours since its last collection.
Each trigger can be limited to some disks. A disk seen for the first time fires nothing: a disk failed for months is not news. The last state of each disk is kept by the integration, so a disk that failed while Gladys or the integration restarted is still announced, once.
Actions
- State of a disk: reads a disk for the next actions: failed or not, its health, name, model, temperature, power-on hours, hours since the collection, the attributes in cause, and a ready-made sentence such as "Samsung SSD 840 (sdd): Failed, 33 °C, collected 5 days ago. UltraDMA CRC Error Count."
- Count the failed disks: counts the disks failed, with a warning, no longer collected and in total, and names the failed ones, for a report.
Both read Scrutiny again when the last reading is more than a minute old.
Scene ideas
- A disk fails → send a Telegram message "Disk {disk} failed: {reason}", and light a lamp in red.
- Every Sunday at 10:00 → Count the failed disks → if failed is above 0, send "{names} failed, {warning} disk(s) with a warning".
- Disk no longer collected → notify "The Scrutiny collector of {disk} stopped {hours} hours ago": a backup NAS that did not wake up, a collector container that crashed.
Troubleshooting
- "Scrutiny does not answer": check the address and the port from the machine of Gladys, and that a firewall does not block it.
- "Scrutiny answers with an error": its database (InfluxDB) is down or full; the Scrutiny dashboard shows the same error.
- "A login is asked": a reverse proxy protects Scrutiny; fill in its user and password (basic authentication only).
- "The address redirects": enter the
https://address directly. - "Certificate refused": a self-signed certificate is not accepted; use
the
http://address on your network. - No disk: no collector sent data yet; run it once
(
docker exec scrutiny scrutiny-collector-metrics runwith the omnibus image). - A disk stays "Not collected yet": it is registered but its collector could not read its S.M.A.R.T. data (a USB enclosure, a RAID controller): see the collector logs.
- The health differs from the Scrutiny dashboard: reload the dashboard; the integration reads the same status threshold, at the chosen interval.
Limits
- The data is the one of the last collection: with the default daily collector, a disk failing at noon shows at the next collection.
- Scrutiny's history, its notifications and the self-test results are not read; the history of each feature is kept by Gladys.
- The sector counters exist only for SATA disks; NVMe and SAS disks show their health, temperature and hours.
- Not affiliated with Scrutiny or its author.
Configuration settings
These are the settings Scrutiny asks for in its configuration screen in Gladys.
| Setting | Type | Required | Description |
|---|---|---|---|
| Your Scrutiny | section | No | Enter the address of the Scrutiny web interface. Then add the disks to follow from the Discover tab: Gladys reads Scrutiny, the Scrutiny collector reads the disks. |
| Scrutiny address | string | Yes | Address and port of the Scrutiny web interface (8080 by default). |
| Proxy user (optional) | string | No | Only if a reverse proxy asks for a login in front of Scrutiny (basic authentication). |
| Proxy password (optional) | secret | No | The password of this user. Scrutiny itself has no login. |
| Options | section | No | |
| Refresh interval | select | Yes | How often Scrutiny is read. Its collector reads the disks once a day by default. |
| Collection too old after | select | Yes | A disk not collected for this long fires "Disk no longer collected": a stopped collector, a disk gone. |
| Language of the names | select | Yes | Language of the feature names (set when a device is added) and of the texts sent to scenes. |
How to install Scrutiny in Gladys
- In Gladys, open Integrations: Scrutiny appears in the catalog, next to the native integrations, with a community badge.
- Click Install. Gladys pulls the Docker image (
ghcr.io/guim31/gladys-scrutiny:1.0.1), starts it in a sandbox isolated from the core, and generates the integration's interface (devices, discovery and configuration). - Open the Configuration screen of the integration, fill in the settings, and save.
- You can also install it directly from its repository URL: https://github.com/guim31/gladys-scrutiny.
Scrutiny requires Gladys >=5.1.0. The catalog inside Gladys refreshes every hour, so a new version becomes available at most one hour after its release.
Not running Gladys yet? It is free and open source: follow the installation guide to get started.
About external integrations
Scrutiny is an external integration: a community integration packaged as a Docker container and published on GitHub, that Gladys installs in one click and runs in a sandbox isolated from its core. It is published and maintained by guim31, not by the Gladys core team.
- Browse all integrations, native and community
- Build and publish your own external integration
- Source code on GitHub — source of this documentation