This is the quickest way to get OOTT running. OOTT is published on Docker Hub as a pre-built image, [`rzuasti/oott`](https://hub.docker.com/repository/docker/rzuasti/oott/general), and the steps below get you to a running service with the smallest possible configuration.
You need somewhere on the host to store the configuration file and the database (OOTT uses SQLite, so the database folder should be on local — not remote — storage):
The user that runs Docker needs read access to the config folder and read/write access to the db folder. You can use any paths you like as long as you keep the volume mappings below in sync.
This is enough to start the service; every other option falls back to a sensible default. See [Full list of configuration options](#full-list-of-configuration-options) for everything you can tune, and the [full sample config](https://github.com/rzuasti/oott/blob/main/examples/sample_oott.toml) for a commented, all-options file.
Create a `docker-compose.yml` at `/docker/oott/docker-compose.yml`:
```yaml
name:oott
services:
oott:
image:rzuasti/oott:latest
volumes:
- /docker/oott/config:/config
- /docker/oott/db:/db
network_mode:host
```
Host networking is required so OOTT can see the multicast and broadcast traffic its scanners rely on (see [Privileges and network ports](#privileges-and-network-ports)).
Start the service from where you placed the compose file:
```bash
docker compose up -d
```
That's it. Check that everything is running with `docker ps`, and follow the logs with `docker logs CONTAINER_ID -f`.
The official OOTT apps for **iOS and Android** are available as a **one-time paid purchase** on the App Store and Google Play. Buying them is the easiest way to get OOTT and directly supports its ongoing development. 💚
OOTT is and will remain **fully open source**. If you'd prefer not to pay, you're free to **build and run the apps yourself** from this repository at no cost, under the terms of the AGPL-3.0. (If you redistribute your own build, see the [Trademark](#trademark) note — it must use its own name and branding, not "OOTT".)
The OOTT app (web, desktop, iOS and Android) talks to the backend exclusively over its REST API on port `3000`. Open the app's settings and point it at your backend's address on the local network over HTTP, for example `http://192.168.1.50:3000`, using the API key you set in `web_server.api_key`.
This is the simplest setup and needs no extra infrastructure, but it only works while your phone is on the same local network as the backend.
> On iOS, the direct HTTP connection only works when you use the backend's **private-range IP address** (`192.168.x.x`, `10.x.x.x`, `172.16–31.x.x`) or a `*.local` (mDNS/Bonjour) name. A custom internal domain (e.g. `oott.mylan.com`) served over plain HTTP is blocked by iOS App Transport Security, even when it resolves to a private IP. Use the IP address directly, or set up HTTPS.
> The first time the iOS app reaches your backend on the local network, iOS shows a one-time "find devices on your local network" prompt. This is expected — tap Allow to continue.
To use a domain name, or to reach the backend from outside the local network, set up TLS as described in [Using HTTPS and domain names with the mobile apps](#using-https-and-domain-names-with-the-mobile-apps).
OOTT comes with a pre-built NixOS flake that you can integrate into your configuration. If you are not using flakes you can use the [flake code](https://github.com/rzuasti/oott/tree/main/nix) as a baseline and write your own derivation.
To integrate the OOTT flake into your config you generally do three things: add OOTT to your inputs, add its module and overlay, then enable and configure the service.
**3. Enable and configure OOTT** — in your `configuration.nix` (or an imported file). Set all options through the service definition; the keys mirror the config file options listed [below](#full-list-of-configuration-options):
The system configuration lives in a single config file, which you can write in TOML, JSON or YAML. With Docker, TOML is recommended — the [full sample config](https://github.com/rzuasti/oott/blob/main/examples/sample_oott.toml) lists every supported option with comments. With the Nix flake, set the same options through the service definition (see above).
Options marked **Required** have no built-in default and must be set in your config file. Everything else falls back to the default shown if you omit it.
|`database.path`|**Required**|Location of the system database. Must be `/db/oott.db` when using the Docker image.|
|`networking.interface`|auto-detected|Network interface to use for scans. Optional — if not set, the first non-loopback connected interface is used automatically.|
|`web_server.api_key`|**Required**|API key the app must present to use the backend.|
|`arp_scanner.enabled`|`true`|Whether to run the ARP scanner. The whole `[arp_scanner]` section is optional; omit it to use the defaults below. Set to `false` to turn it off.|
|`arp_scanner.wait_between_scans`|`30m`|Time to wait between each network scan (you can express it in seconds, minutes, hours, etc. as a suffix - for example: 30s, 10m, 1h)|
|`arp_scanner.sender_timeout`|`1m`|If the ARP sender process takes longer than this it will be stopped (for a class C network - 254 IPs - it should take less than a minute)|
|`arp_scanner.scan_duration`|`10m`|How long to wait for response packets on each scan (5m to 10m is a good timeframe for a class B or C network)|
|`mdns_scanner.probe_timeout`|`2s`|When an mDNS-discovered IP is not in the OS ARP cache, how long to wait for a targeted ARP probe reply to resolve its MAC address|
|`ssdp_scanner.probe_timeout`|`2s`|When an SSDP/UPnP-discovered IP is not in the OS ARP cache, how long to wait for a targeted ARP probe reply to resolve its MAC address|
|`dhcp_scanner.enabled`|`true`|Whether to run the DHCP scanner. Set to `false` to turn it off.|
|`snmp_scanner.enabled`|`false`|Whether to run the SNMP scanner. The whole `[snmp_scanner]` section is optional and the scanner stays off unless you add it; once the section is present it defaults to `true`.|
|`snmp_scanner.target`|**Required**|SNMP agent to poll, as `host:port` (e.g. your gateway: `192.168.1.1:161`). The scanner reads the agent's ARP table over SNMPv2c — no local probing. Required when the `[snmp_scanner]` section is present.|
|`snmp_scanner.community`|**Required**|SNMPv2c read-only community string. Use a read-only community and never commit a real secret. Required when the `[snmp_scanner]` section is present.|
|`snmp_scanner.wait_between_scans`|`10m`|Time to wait between polls. Keep it well under the agent's ARP cache timeout so active devices aren't missed.|
|`notifications.notify_when_not_seen_for`|`1w`|Send a notification if a device comes back online after not being seen for this timeframe (you can use hours, weeks, etc.)|
|`notifications.pushover.token`|**Required when `method` is `pushover`**|Your pushover token goes here, just copy&paste from their website after creating the app. The whole `[notifications.pushover]` section may be omitted when `method` is anything other than `pushover`|
|`notifications.pushover.user_key`|**Required when `method` is `pushover`**|User key goes here, this is the account wide code for pushover. The whole `[notifications.pushover]` section may be omitted when `method` is anything other than `pushover`|
|`retention.window`|`365d`|How long to retain device events and notifications. Records older than this are purged daily. Accepts duration strings (e.g. `90d`, `1y`, `6m`).|
|`device_events.deduplication_window`|`1m`|Suppress duplicate device events: if the same scanner sees the same device (same MAC and IPv4) again within this window, only one event is recorded. Accepts duration strings (e.g. `30s`, `1m`, `5m`).|
### Using HTTPS and domain names with the mobile apps
To reach the backend through a domain name, or from outside the local network, put it behind a reverse proxy (nginx, Caddy, Traefik, …) that terminates TLS with a **valid certificate from a trusted CA** (for example [Let's Encrypt](https://letsencrypt.org/)), then point the app at the HTTPS URL (e.g. `https://oott.example.com`). This works on every platform — including iOS — with no further configuration.
Connect using the **domain name the certificate is issued for**, not an IP address. Self-signed certificates are not supported unless they are manually trusted on the device.
OOTT does not provide built-in user accounts or permission management. Access to the backend is gated only by the `web_server.api_key` you configure.
If you want proper user authentication (login pages, multiple users, single sign-on, etc.), add an authentication layer in front of OOTT — for example [TinyAuth](https://github.com/steveiliop56/tinyauth), [Authentik](https://goauthentik.io/) or a similar reverse-proxy authentication solution.
> [!IMPORTANT]
> Do **not** apply that authentication layer to the `/api` URIs. The mobile apps and web front-end talk to the backend over `/api` using the API key.
OOTT stores a timestamped event for every device *sighting*, so the database grows with the number of devices on your network and how long you keep history.
**2. Enabled scanners** — each scanner you turn off removes its share. The three passive scanners (mDNS/SSDP/DHCP) generate most of the events; disabling them leaves just the ARP baseline (≈1.9 MB/device/year).
**3. ARP scan interval** (`arp_scanner.wait_between_scans`) — a longer interval means fewer ARP events. This only affects the ARP share, so its overall impact is modest while the passive scanners run.
> The deduplication window (`device_events.deduplication_window`, default `1m`) caps how often a single device can be recorded — at most one event per scanner per minute. It's a safety limit for chatty devices, not a routine tuning knob.
The mDNS (UDP `5353`), SSDP (UDP `1900`) and DHCP (UDP `67`) ports are fixed by their respective protocols and **cannot be changed**. They must be available on the server where OOTT runs.
OOTT binds these sockets with address/port reuse, so it can run alongside other responders already listening on them (for example `avahi` on `5353`, `minidlna` on `1900`, or a DHCP server/relay on `67`). However, the ports must not be blocked by a host firewall, and the corresponding multicast/broadcast traffic must be allowed to reach the host — otherwise the scanners will not discover any devices.
> [!IMPORTANT]
> OOTT needs elevated privileges: the ARP scanner requires raw-socket access, and port `67` is a privileged port. The pre-built Docker image and NixOS module already run with what they need.
>
> Under Docker, the scanners require **host networking** (or an equivalent setup that exposes the host's multicast and broadcast traffic to the container), as shown in the [Docker installation](#simple-installation-with-docker) above.
"OOTT", the OOTT name, and the OOTT logo are trademarks of Ricardo Zuasti and are **not** licensed under the AGPL-3.0. The AGPL-3.0 grants you full rights to the *source code* — to use, modify, build, run, and redistribute it — but it does **not** grant any right to use the OOTT name, branding, or logo.
If you build or distribute your own version of this software, you must do so under a **different name and branding** and must not present it as the official OOTT app. This keeps users from confusing third-party builds with the official, supported release.
