Environment: Ignition 8.3.7, MQTT Transmission 5.0.4, Sparkplug B to a third-party MQTT broker. OPC UA source is a third-party server; Ignition is the OPC UA client (subscription based).
Problem: When an OPC UA node becomes unavailable (source-server update, or the tag is deleted upstream), Ignition reports Bad_NodeIdUnknown and refreshes the tag’s timestamp every ~10 s. This is probably wanted behavior for a failed-item retry loop, i have not found the setting which controls this. Each refresh produces a DDATA with is_null=true and Quality 0xC0000300. That’s ~6 null publishes per minute per tag, across every transmitter subscribed to it, for as long as the outage lasts.
reproduction: Take a working subscribed OPC tag and change its Item Path to a non-existing node (simulating a deletion). Observe DDATAs has null values for the tag every 10 second.
Question: Is there a MQTT Transmission-side setting or supported scripting to skip publishing metrics when the tag is in “error”, without a Transmitter refresh on every transition? Or any other suggestion to handle this?
Pictures:
Decoded DDATA
Ignition tag that has the “null” values sendt:
This error tag is of course a transient problem and will be fixed, but we don’t want to automatically delete tags.

