OPC UA unique ID

Question

I'm trying to build a OPC UA client application.
I'd like to be able to identify a UA node uniquely in the OPC tree.
I know that in OPC DA, a standard node id is a string with a '.' as a delimeter that I can use in order to identify a node.

In OPC UA, the node ID doesn't have to be a string, but I'd still like to be able to build a unique string that maps to a particular node.
I'm thinking about basing it on the the nodes names. e.g.: Demo.MyNode.MyValue.
but I'm afraid that the node name can contain characters such as "." and this will make my IDs not unique.

Is there a character I can use as a delimeter?
Is there a better way to represent the node ID as a string (including its path)?

Answer 1

OPC-UA offers the concept of a unique "BrowsePath" for each and every node, and a client could opt to store BrowsePaths instead of NodeIds, and then upon startup call the TranslateBrowsePathsToNodeIds service.

In fact, I believe this may be the intended behavior, as there's no requirement that a server use the same NodeId for any given Node after restarting, even if in practice that's how it's done.

I was wrong about NodeId being allowed to change. The spec says: "A Server shall persist the NodeId of a Node, that is, it shall not generate new NodeIds when rebooting."

I now believe its best to store NodeIds and only use BrowsePaths to aid in programming against type definitions.

Answer 2

One of the features of OPC UA is that the server can offer different menu trees to different users. It may not matter for your client, since any given user will only see the one tree, and the BrowsePath will be unique for that user.

In v1.03 of Part 3 of the OPC UA spec, "OPC UA Part 3 - Address Space Model 1.03 Specification.pdf", section 5.2.2 says a server should not change a node's NodeId when it reboots. (The spec is available from the OPC Foundation at https://opcfoundation.org. You can register and download it for free.)

Of course, some UA servers might not maintain their NodeIDs across a reboot. Which is another reason to use Kevin's suggestion to use the BrowsePath to make a unique string for each node. The string can make it clearer to the user which node they're accessing. Good idea!