Remote Tag Providers in Enterprise Systems - Does Edge break the cardinal rule?

I’ve been actively interested in utilizing the EAM capability in Ignition in an Enterprise Hub and Spoke Architecture. At the 2018 ICC, several sessions covered enterprise topics, and after listening to the one Travis Cox presented (well done, Travis), I was relieved to hear many good points on pitfalls to avoid. One such recommendation - rename the default real time tag provider to be unique - was something we discovered on our own, along with many of the others.

So, Why is this important? We have been looking at using EAM and a central system to create and maintain tags, projects, and resources for all of our sites. For tag providers, running a project on a central system that will also run on a remote system requires the tag provider be unique across all gateways (no overlap of tags) so that the tag path in your project will still resolve correctly regardless of the gateway it is run on, whether it be a Remote Tag Provider or a local one.

I guess I missed a few things in my test, and I think IA may have broke their own cardinal rule with Edge (or I haven’t figured out how to use it correctly yet… most likely!). In Edge, the real time tag provider is “edge” and cannot be renamed. The other thing I found is that the project name is also fixed, meaning importing and managing the project on a central system also doesn’t seem possible. Am I missing something?

Note: We have multiple Edge Panel Enterprise systems installed.

Crickets? Several months and no reply probobly means you are the only one pondering this delema.

im new to the EAM and Edge architecture and I have a couple of questions that perhaps you can answer.

  1. With EAM on central server and Edge enterprise on the remote server, is it possible to have a project from the main server be published to the edge panel whereupon the edge panel will serve as the lical client fallback?

  2. Where do the tags reside for the project on the main server. do the exist on the remote edge panel? Im conserned about tag addressing being broken in the event of a fall back.

  3. Whats the point of a local client fall back if the tags are all on the edge anyway?

  4. What if the main project has features that do not exist on the edge such as db tables, reports, tranactuon groups, etc.

  5. Is it possible to export a project from the main server such as windows and tags and import them into the edge designer thus expiditing the development process for a customer that wants to migrate to a distributed architecture?

Nothing official here, but:

1: No.
2,3,4: Not applicable without fallback.
5: Probably, if you avoid all of the features Edge doesn’t have.

You might be wishing for more than Edge is intended to do.

  1. If you’re looking at it from a “project” standpoint, Ignition Edge != Ignition. So depending on exactly what you want the system to do, Edge might not be the solution but a Single Vision Ignition Client w/PLC Drivers/Tag History…etc might be.

  2. Tags can exist as remote tags coming from an Edge device or it could be on the main Ignition gateway. For fall back to work, tags would have to be on the Edge device as falling back to Edge would still require the tags to work properly. Also, if you need to take advantage of store and forward, then Edge needs to own the tags.

  3. The point of local client fall back is to ensure you still have an interface local to the equipment should you loose connectivity to the primary Ignition server. Say IT closes a port on a switch, or performs an upgrade/maintenance activity, Local client fallback should have comms that can directly access the local PLC, thus operators can still have basic control of the machine. It’s also possible that the local machine has a third party HMI like a PanelView they could use. So Edge might be added protection for ensuring you don’t loose data (store and forward).

  4. Well, Edge is different than a standard Ignition application. So you need to understand the use case for Edge. It isn’t to replicate standard Ignition, because what’s the point then? Just get a single client license of standard Ignition .

  5. Portions of it yes, but it comes down to what you need Edge to do versus a standard Ignition application.

The point I was attempting to understand is how to achieve using Edge Enterprise in the architecture that is described in the hub and spoke design. One of the advantages (local client fallback is as Paullys50 indicated, meant to provide local control in the event a connection with the Central Ignition Gateway is lost) I saw around using Edge was to allow multiple client connections to what is otherwise a single client connection (utilizing the full Ignition install). This requires development of the project on the full install, and another project built specifically for the Edge system (I think). What isn’t clear to me is how this setup is done and how users should connect.

Your point in 2 is valid, and on a single Edge to Full installation, the Remote Tag Provider can be the same as on the Edge system (Edge - default and unchangeable), and projects developed on the full install (will also be called Edge, but can have a different title) can be published from full to edge. If you have more than one Edge system connected via the gateway network, the Remote Tag Provider now has to be unique, and projects as well. Thus, the single project now must be two (to handle the different tag provider names) as well as the project name.

On point 4, this is exactly a concern. Edge serves a purpose, and I found that if all you want is data collection (Edge Enterprise or Edge MQTT), then it is perfect for that application. If you want an HMI built specifically with no thought of integrating to a larger distributed system, then Edge Panel will suffice with your needs. However, if you want a local project with easier management from a central Ignition instance, then you are better off with the single client full Ignition instance. We learned the hard way and had to upgrade to full Ignition instances (IA makes this very easy by the way, so the initial investment isn’t a loss). You will have to ask for the single client license of Vision, and is comparable to the Edge Panel. You gain some nice features, but you lose some too for the price (historian mainly).

Hope that helps.

1 Like