DeviceChain
English
Download
Integrations · Sparkplug B

Sparkplug B host application for DeviceChain

Many industrial and building-automation fleets already publish telemetry as Sparkplug B over an MQTT broker they operate. DeviceChain can ingest from those networks without asking the devices to change anything. It joins the environment as a Host Application and translates the edge traffic into the same events every other transport produces.

Sparkplug B

Connecting out to your broker

With plain MQTT, DeviceChain is the broker. Sparkplug ingestion works the other way around: the optional Sparkplug service connects out to your broker as a client, subscribes to the groups you configure and follows the Sparkplug session protocol. It announces itself with the Sparkplug STATE message so edge nodes know a consumer is online.

Each source is configured for one tenant, with a broker URL, credentials supplied as a projected secret and the groups to subscribe to. The tenant is fixed by which broker the message came in on and is never read from the topic, so one tenant's edge network cannot be mistaken for another's.

Sparkplug B

What it understands

DeviceChain runs the session state machine: it tracks each node's metric aliases and message sequence, and detects a gap or a missed birth, and it can ask a node to rebirth to resynchronize. A group, node and device become the external id of a DeviceChain device. With auto-registration enabled for a source, a device is created the first time it is seen.

A node or device BIRTH marks the device online and a DEATH marks it offline, immediately and explicitly, so presence is authoritative rather than inferred from a timeout. Decoded metrics flow into the normal pipeline, so history, live state, dashboards and rules treat Sparkplug telemetry like any other.

Sparkplug B

Commands

Commands down to a node or device over the same Host Application session are opt-in and off by default. When a source enables them, the console can declare standard commands that write one metric, ask a node to rebirth, or reboot a node under extra conditions. Sparkplug has no command acknowledgement, so the platform waits for the node to report the requested value.

Sparkplug B

High availability

Only one replica of the Sparkplug service connects to a given broker at a time, enforced by a fenced ownership lease, because a Host owns per-node session state. A replacement pod acquires the lease and re-establishes the session before it serves.

Not yet

What it does not do.

Sparkplug-B ingestion → · Transport matrix → · All integrations →