Hi,

The unique identifier for a node is the NodeId Attribute. I’m not exactly sure what you mean reading by “their name” as you do not operate on names. Nodes have the BrowseName and DisplayName Attributes, which is usually displayed instead of the NodeId in e.g. UIs as a Tree view, however all the operations are done based on the NodeId attribute (i.e. BrowseNames and DisplayNames are not unique, NodeId is). You can read OPC UA Specification (v.1.03) Part 3 for the address space concepts.

Every node is a target of a Hierarchical Reference from another node, with the top being the Root node, with numeric Id 84 of the standard namespace, under which e.g. the Objects folder is (usually the Root is not displayed). If you know nothing about the server beforehand, that is all you can expect, however usually in a client application you know the NodeIds directly based on e.g. a configuration file or e.g. have the NodeId under which the relevant nodes are which you can then Browse.

Assuming you know the starting point and a set of BrowseNames you can use the TranslateBrowsePathsToNodeIds service call to get the NodeIds relative to other NodeId. You can call this in the SDK via client.getAddressSpace().translateBrowsePathsToNodeIds(…). Please note that the like NodeIds, the BrowseNames do have a namespace index, which must be correct in order the call to work. For more information read the javadoc of the SDK method and UA Specification (v.1.03) Part 4 section 5.8.4.

– Bjarne