UNAS Monitor
A NAS earns its keep by being forgettable. The shares mount, the backups run, and nothing asks for your attention until the morning a disk has already gone. UNAS Monitor puts the inside of the box on your dashboard so that morning comes with some warning: the state of the array, and for every disk its temperature, its health score, and the day its first bad sector appears.
It reads your Ubiquiti UNAS directly on the local network. The array becomes one device, and every disk becomes its own device, which matters more than it sounds. Homey then keeps a separate Insights curve for each bay, so you can see that bay 3 has been running two degrees warmer than its neighbours since July, and a Flow can name the disk that worries you instead of saying “something is wrong with the NAS”.
What the array reports
| Capability | What it tells you |
|---|---|
| Array state | Fully operational, or what it has degraded to |
| Array usage | How full it is, in per cent |
| Free space | What is left, in TB |
| Disks | How many the array holds |
| Temperature | Of the appliance itself |
| Processor load | In per cent |
| Memory used | In per cent |
| Read and Write | Disk throughput, in MB/s |
| Network in and Network out | Link throughput, in MB/s |
| Array degraded | Alarm, raised the moment the array stops being fully operational |
| Array almost full | Alarm, raised above the threshold you set |
What each disk reports
| Capability | What it tells you |
|---|---|
| Disk state | Optimal, or whatever the console now calls it |
| Temperature | In degrees, per bay |
| Health score | As the console computes it |
| Power-on hours | How long this disk has been spinning |
| Bad sectors | How many it has lost |
| Read and Write | Throughput for this disk alone, in MB/s |
| Disk not optimal | Alarm, on any of the four grounds below |
The disk alarm rises on four separate grounds rather than one: a state that is no longer optimal, a temperature above your threshold, the very first bad sector, or a health score that has started to slip. The first bad sector is the one worth waking up for, because a disk rarely loses exactly one.
Flow cards
Six triggers and four conditions. No actions: the app watches, it does not touch your storage.
Triggers:
- The array leaves the fully operational state
- Array usage rises above a threshold you set on the card
- The processor rises above a threshold you set on the card
- The disk leaves the optimal state, for one named disk
- The disk rises above a temperature you set on the card, for one named disk
- A bad sector appears, for one named disk
Conditions:
- The array is, or is not, fully operational
- Array usage is, or is not, above a threshold
- The disk is, or is not, optimal
- The disk is, or is not, hotter than a temperature you set
Thresholds live on the Flow card rather than on the device, so one disk can feed three Flows set at forty-five, fifty and fifty-five degrees without any of them fighting over a single setting.
Widgets
Two of them, Disk array and Disks. Each takes a background: Night, Slate or Plum.
Settings
| Setting | Where it lives | Default |
|---|---|---|
| Refresh interval | On the array | 300 seconds |
| Array usage threshold | On the array | 85 per cent |
| Disk temperature threshold | On each disk | 50 degrees |
The array polls once for everybody. A four bay machine makes three calls per cycle rather than fifteen, because the disks read the same answer the array has just fetched.
An API key, and nothing else
No account, no cloud, no port forwarding. Create a key on your console under Control Plane, Integrations, Create New API Key, then pair the array with the console address and that key. The disks are discovered from the array, so you never type anything twice.
If you rotate the key later, use Repair on the array rather than deleting it. The device keeps its identity and its Flows, and the new key replaces the old one in place. The key is stored with the array and is deleted along with it.
Everything stays on your network. The console sits on your LAN, the key is only good for that console, and nothing is sent anywhere.
Which models
Written for the whole range rather than for one box. Ubiquiti sells two, four, seven and eight bay models, and nothing is assumed about how many pools you have created or how many disks sit in each: the app reads every pool and attaches its own disks to it. Empty bays are recognised as empty rather than counted as failed disks.
Languages
English, French, German, Spanish, Italian, Dutch, Polish, Danish.
Reporting a problem
Post here with your Homey firmware version, the app version, your UNAS model, and how many bays are populated. A screenshot of the device page helps more than a description. If pairing is what fails, the error message is the useful part: it says whether the console could not be reached at all or whether the key was refused, and those two send you to very different places.
Feedback and testers welcome.
Have fun 
