“Are Nodes ID’s unique in the whole OPCUA server?”
NodeIds are unique in the server.

“Is it enough to specify the NodeID in the read method?”
The NodeId will specify the Node for which to read so yes. If you want to read the Value Attribute then UaClient.readValue(nodeId) is enouhg. If you want read other Attributes then you need to use UaClient.readAttribute. There are overloads which allows spefying max age and index range too.

“How to specify the fully qualified tag path of a node?”
Usually you know the NodeId of the node directly. If the server supports it, you can use TranslateBrowsePathsToNodeIds Service from UaClient.getAddressSpace().translateBrowsePathsToNodeIds, but I would say it is easier to use the NodeIds directly..

“How to read group of tags?”
For Value Attributes, e.g. by UaClient.readValues(NodeId[]). For other Attributes, construct ReadValueIds and pass them to UaClient.read

– Bjarne

“ns=2;s=…” means it is a NodeId with NamespaceIndex 2, Identifier type is a String (the ‘s’), and the part after “s=” is the value part of the NodeId, i.e. it is not a path, just an id for a node. For example the “ns=2;s=[default]New Tag1” would be created as

.

By path I tought you meant e.g. “(Root)/Objects/MyObjects/MyDevice/MyLevel” (The MyLevel sample node in the SampleConsoleServer). The Translate.. Service call uses BrowseNames (BrowseName is an attribute of a node, it also has a namespace index in addition to the string part), for visual usually the DisplayName attribute is used. Anyway the service is on the server side so it’s functionality depends on the server. There are not many cases where knowing the path is of any use if you know the NodeId directly (e.g. you have Browsed the address space).

If you don’t have them yet, I would suggest you download the specifications from https://opcfoundation.org/developer-tools/specifications-unified-architecture (you need to register, but otherwise it is free). The part 4 lists all the Service calls you can make to the server (note that some of them are optional so it depends on the server). It will also make understanding some parts of the SDK easier.

– Bjarne

Probably better to read from the spec, but essentially the same thing as in a programming language namespaces (or packages in java): to avoid naming conflicts.

The 0 namespace index is always the one defined by the specification, i.e. the standard namespace. 1 is for server itself (e.g. for some diagnostic nodes and such). After that it depends on the server. The NodeId data types is defined in the part 3 (section 8.2), there are 4 data types. Usually it is numeric or string (some diagnostic nodes use GUIDs). The namespace index is defined as UInt16 so there is theoretically 65535 possible namespaces in the server.

(some more explanation below, depending what you do this might not be that relevant for now)
I guess I should also mention that the namespace index points to the Objects/Server/NamespaceArray’s values. It contains the NamespaceURI for that namespace. See Part 3, section 8.2.2. The basic idea is that data transfer between client and the server uses the namespaceindex. Since it is a number it is shorter (smaller in binary) to transfer than e.g. a 20 character URI. Since the NodeIds are used a lot in the communication this makes it more efficient. In practice this means the client should check the NamespaceArray of the server to see which index matches which URI (because the order is not fixed beyond 0 and 1). UaClient fetches this information for you then it is connected, see UaClient.getNamespaceTable().

– Bjarne