Hi, I have an architecture that contains one central Ignition Gateway which is connected to multiple edge gateways. So I am sending data to the Ignition Gateway using MQTT from Edge. So I have a few doubts in this case:
- Is there any way we can maintain the same tag structure from Edge to the Central Gateway so that we can reuse the Objects, symbols & Windows we created in the Edge Gateway Project to the Central Ignition Gateway project?
- I want to publish the data from Edge to 2 brokers, both have an additional broker that is used for redundancy(so total 4 brokers). So in this case can I subscribe from Central Ignition Gateway using 2 brokers with the same namespace
For #1, I always use reference tags in real-time tag providers to reference the tags in the MQTT Engine provider. This way you can structure them however you'd like in as many different tag providers as you'd like. This way if each edge unit maps to a unique real-time tag provider you can, or use a single tag provider and make a folder for each edge unit and structure the tags under each to match the edge structure. To make life easy, take your UDTs from the edge and rework them to be reference tags instead of OPC or memory tags.
For #2 ID have to look into it again, but I don't think MQTT Engine can subscribe to 2 separate brokers for redundancy. For redundancy, you'll probably need to set up an MQTT cluster that Engine and Transmission then have both servers in their configurations and can connect to either of them, but only ever 1 at a time. I know EMQx is capable of clustering, but I haven't used other brokers to know their capabilities.
To be honest, this sounds a lot like the exact application you would use the gateway network, not MQTT.
The issue is in my architecture I have a Regional Control Center with a redundant Edge pair and a Backup Control Center with a redundant Edge Pair. Which needs to publish the same data to the National control centre. At the national level 4 datasource will be available but only 1 we will use & it needs to switch over automatically.
What you'll need to do in that case is have the data published to 2 separate "folders" or sites, then at the national level, look at both and use the data from the one with good quality for every tag. This will give you granular redundancy. If you don't want to do it at the tag level, you could do it on a site level, but then which tags do you monitor and what if a handful of tags go bad, but they're good at the backup.