Reference tag resource cost vs. another tag

What is the cost savings (if any) in terms of system resources of a Reference tag vs. another instance of an opc tag?

CPU?
Memory?
Network?
PLC?

For just a handful of duplicate tags I would imagine the cost difference is negligible. But if there is any cost savings, how does it scale when considering 100's or 1000's of tags? And is there some known point of marked difference.

Cheers!

We are also interested on that topic, have you found an answer?
Thanks a lot!

For OPC tags, I would just make another. Ignition will optimize away the extra traffic as long as the subscription pace is the same. Then nothing breaks if the original tag is deleted.

I would consider this worth it even if the extra OPC tags took more resources than a reference tag. (But I doubt the resource difference is significant.)

What's the use case where you would be considering adding a reference tag vs duplicate OPC tag?

I have a case where we have several process values that each have a UDT with limit and alarm controls, but the process value also controls a PID or similar.

We often just display the process values grouped by their location on the machine/product. We also have templates/popups for viewing/controlling the PIDs and want to display the process value associated with the shown PID.

Each PID has its own folder that contains tags linked to the different control values in the PLC as well as the CV and PV.

Currently I just make a the PV tag in the PID folder a reference tag that points to the process value in its UDT. This allows me to have a single UDT instance per process value while still associating the process value with the control items that utilize it.

We have OPC tags that retrieves the value, but we forward this signal to other layers, so we have a folder with all signals that are forwarded (for clarity), and we have the doubt if it is worth to duplicate the signal or have a reference to the original OPC.

Reference tag seems appropriate here.

A duplicate OPC tag has more overhead and cost, as it means a duplicate monitored item on the server and a duplicate item being polled depending on the server and driver implementation.

1 Like

Thanks a lot Kevin :muscle: