Step-by-Step API Workflow Guides

A single click in the Litmus UI often triggers a chain of API calls behind the scenes. Each guide below documents one of those chains end to end: the exact call order, the request payloads, which IDs to capture from each response, and the pitfalls the UI quietly handles for you. Use a workflow guide when your task is a verb ("create a device", "deploy an app"); use the per-product endpoint reference when your task is a noun ("list devices", "get a tag value").

Litmus Edge workflows

Create a device with tags

Pick a driver, create and start a DeviceHub device, then add the tags to poll.

5 steps

Browse for tags in bulk

Crawl a device's full address space (OPC UA nodes, S7 program tree, ...), pick items into a cart, and create them all as tags in one commit.

7 steps

Build an analytics pipeline

Wire source, transform, and sink processors into a working pipeline, step by step or as a single importable JSON document. Includes OEE, anomaly detection, and SPC recipes.

5 steps

Create an integration instance

Send tag data to Kafka, Azure IoT, AWS IoT, InfluxDB, S3, and other providers: read the provider schema, create the instance, subscribe tags, start it.

4 steps

Configure the OPC UA server

Expose edge tags upstream to SCADA or MES systems through the embedded OPC UA server.

4 steps

Deploy an edge application

Install and launch a containerized application from the marketplace onto the edge.

4 steps

Apply or upload a configuration template

Download a device configuration snapshot and apply it to another edge using the resumable upload session flow.

3 steps

Restore a system backup

Upload and restore a full edge backup.

3 steps

Upload a custom CA certificate

Install a trusted certificate so the edge can talk to internal services over TLS.

2 steps

Upload an analytics or TensorFlow model

Push an AI/ML model to the edge Analytics engine for local inference.

3 steps

Authenticate a NATS / DataHub client

Open the device's internal NATS bus to an external client: enable the proxy, create an access account, grant topics, then connect with the four LE-specific client settings.

5 steps + connect

Litmus Edge Manager workflows

Provision a new edge device (zero-touch)

Pre-register an edge so it self-activates and joins the fleet on first boot.

5 steps

Deploy an application to an edge device

Push an application to a remote edge through LEM's asynchronous task system, polling until done.

5 steps

Back up and restore an edge from LEM

Take a fleet-managed backup of one edge and restore it to the same or a replacement device.

5 steps

Create a project alert

Define a trigger condition and the notification action it fires across a project.

4 steps

Upload an AI/ML model to LEM

Stage a model in LEM (S3-backed) for distribution to edges in the fleet.

3 steps

Litmus Unify workflows

Provision a UNS MQTT account

Create a broker account with credentials and publish/subscribe rules for a new machine or application.

5 steps

Every guide is also runnable interactively: browse all 56 examples in the API collection.

The resumable upload pattern

Several Litmus Edge workflows (templates, backups, certificates, models) share one upload pattern worth knowing before you start:

  1. Clear: a DELETE removes any stale upload session left by an interrupted attempt.
  2. Create session: a POST declares the file size and returns a session ID.
  3. Upload: the file bytes are streamed to the session endpoint, resumable by offset, then the action (apply, restore, install) is committed.

If an upload was interrupted, the half-finished session persists on the server. That is why these workflows begin with the clear step.