Quickstart

This page walks through a first session on a freshly installed gateway, in the order the dashboard's navigation is built: connect, make tags, raise alarms, show it on a screen. It assumes the gateway is running at http://localhost:8080 as described in Installation.

How the dashboard is laid out

The left rail switches between areas, and the panel beside it lists the pages of the area you are in. Which areas you see depends on your role; the admin sees all of them.

Area	Pages
Home	Overview, Watchlist, Activity
Connections	All connections, Devices (one entry per installed driver, OPC UA and MQTT), Databases
Tags	Tags and types, then under History: Historian, Reports, Replay
Alarms	Active, History, Pipelines
Automation	Cron jobs (while the module is installed), Event streams, Script console
Screens	Your start screens, All screens, Editor
Gateway	Diagnostics, Clients, System (Gateway settings, Modules, Project, Modes, Storage, Backup & restore), Security (Users, Sign-in providers, API keys, Audit), Logs
Settings	Appearance, Sidebar, Logo

Connections, Automation and the Editor are shown to engineers and admins; Gateway only to admins. The clock in the top bar starts a replay.

1. Create the first account

Open http://localhost:8080. A gateway with no users shows Set up this gateway with the line "This gateway has no users yet. The first account becomes its administrator."

1. Fill in Your name, Email and Password.
2. Click Create account.

The dashboard calls POST /api/auth/sign-up/email, which succeeds only for the very first user, then signs you in. From then on the page shows Sign in instead, public sign-up answers 403, and only an admin creates accounts, under Gateway > Security > Users. The default role for a new user is viewer; roles are explained in Users and roles.

If an identity provider has been configured, the sign-in page also offers Sign in with that provider. There is no password reset: the page says "Forgotten your password? Ask your administrator for help signing in."

warning

Do this before the gateway is reachable from outside. The first account is the admin, whoever creates it. See Installation.

2. Make a memory tag and watch it live

A memory tag holds a value the gateway keeps itself, with no device behind it. It is the quickest way to see live data move.

1. Open Tags > Tags and types. The page has three groups: the gateway's own @gateway tags, your tag types under @types, and your tags.
2. On the tags group, or on a folder, open the menu and choose Add tag. When several kinds can be made there the menu lists Atomic tag and UDT instance; pick Atomic tag.
3. The editor opens in a panel beside the tree. Give the tag a name and set Kind to Memory ("A value kept by the gateway").
4. Click Create tag.

The tag's row shows its value, quality and timestamp. Type a value in the editor's value control and click Write: the row updates as the gateway publishes the value. Open the same page in a second browser tab and write again; both tabs move, because every tab holds a /sync WebSocket and receives each accepted value.

Writing needs the operator role or higher; the admin has it. The editor also offers scaling, a deadband, documentation and, further down, the tag's alarms and history. Tags explains every kind.

3. Connect a device

Driver modules add a page each under Connections > Devices. With the full image both OPC UA and MQTT are there. Only the OPC UA driver has been run against a real controller (a Siemens S7-1500), and MQTT against a Mosquitto broker; treat anything else as untried.

Add the connection

1. Open Connections > OPC UA (or MQTT). The empty page says "Add your server's address, then tags can read its nodes." (for MQTT: "Add your broker's address, then tags can subscribe to its topics.").
2. Click Add server (MQTT: Add broker). A form opens in the side panel with a Connection section: Name, Address (opc.tcp://host:4840, or an mqtt://, mqtts://, ws:// or wss:// URL) and an Enabled switch ("The gateway connects and keeps trying.").
3. For MQTT, the Sign-in section takes User name and Password ("Leave these empty if the broker lets anyone connect."), Client ID ("Optional. Leave empty and the gateway picks one.") and under Delivery a Default QoS. OPC UA has a Security section for the policy and user credentials.
4. Click Add server / Add broker. The connection appears in the list with its state.

Each row has Test the connection, Edit and Delete. A test asks the driver to open the connection once and reports the result without changing anything. The driver pages are documented in OPC UA and MQTT.

Browse and create tags

1. Click the connection to open it. The detail page shows the Driver figures (Tags subscribed; for MQTT also Topics seen and Last message), Tags with a problem now and the Settings.
2. Click Browse the address space (MQTT: Browse topics). OPC UA shows the server's node tree; MQTT shows the topics it has seen, and a field to Listen on a topic filter for more.
3. Select a node or topic. A live value is read for the selected node. Click New tag to make a tag bound to it; the tag editor opens with the binding filled in. For an OPC UA structure, Create a UDT type from this structure's fields builds a tag type instead.

The dashboard creates the tag through the gateway's own tag API; the driver only supplies the binding. The tag then appears under Tags and types like any other, with the connection's quality.

4. Add an alarm to a tag

An alarm belongs to one tag and watches every value the gateway accepts for it.

1. Open the tag in Tags and types and scroll the editor to its alarms section.
2. Click Add alarm. Give it a name, a severity (1 to 1000, default 500), a condition (greater, less, equal or notEqual), a setpoint, and optionally a deadband and an on-delay.
3. Save. Write a value past the setpoint to your memory tag and the alarm goes active.

Alarms > Active lists what is outstanding, most severe first, and lets you acknowledge. Alarms > History keeps every transition. The states, quality rules and acknowledgement rules are in Alarms.

5. Draw a first screen and open it

Screens (the API calls them views) are drawn in the Editor and run under Screens.

1. Open Screens > Editor and click New screen. Name it; it is created in the selected folder, or at the top level, and opened in the designer. The same page offers Create the example plant, which builds a set of example screens and tags to look at.
2. Drop a widget on the canvas and bind it to your memory tag by its path. Widgets and bindings are described in Widgets and The designer.
3. Save. The toolbar's open button runs the saved screen as operators see it ("Save first: operators see the saved screen.").

Running screens are listed under Screens > All screens; each row opens inside the dashboard. A viewer sees values but no commands; operators can write. See Running screens.

6. Set a start screen

A start screen is a named entry that opens one of your screens and appears in the Screens area of the rail for the people who may see it.

1. In Screens > Editor, click the star on a screen's row. That creates a start screen entry with the screen's name.
2. Pinned screens in the editor's side panel lists them: "Who sees each; the first one someone can see is their start screen."

Under Screens > All screens the start screens come first, marked Default. Several start screens can point at one screen, each with its own name and audience. Start screens has the details.

Where to go next

• Configuration: ports, addresses, retention and the data directory.
• Concepts: tags, tag types, alarms, history and replay, security.
• Modules: installing, starting and stopping drivers; the historian; cron jobs.
• Scripting: expression tags, screen scripts and the system.* API.
• Gateway administration: users, API keys, backups, logs.
• Examples: an S7 over OPC UA, an MQTT sensor, a motor type, an alarm to a webhook.