The ConnectionConfigurationSet can contain a PubSubConfiguration, which has a datatype of PubSubConfiguration2DataType. The PubSubConfiguration2DataType is a complex structure that can include Datasets (published or subscribed) that would include fields that are of NodeId DataType. NodeIds contain a NamespaceIndex. The NamespaceIndex is an index into a NamespaceArray that contains the URI that describes the owner of the Node. On a Server, this NamespaceArray is in the Server Object and is used to resolve all nodes on the Server. But in this case, the PublishedDataSet or SubscribedDataSet describes Variables that are on the Publisher or Subscriber Server, which is very likely to be a different Server. Depending on how the ConnectionConfigurationSet and PubSubConfiguration were generated, the Namespaces may not exist on the ConnectionManager Server.

This document describes how these NodeIds are to be resolved to the correct NamespaceIndex on the target Server before the structure is transmitted as part of an EstablishConnections (see 6.16.3.2). No issues exist in using the structure to set up a Connection, but if the configuration of the ConnectionManager Server is saved as a UANodeSet XML File or if the ConnectionConfigurationsSet is generated in an engineering tool and loaded into the ConnectionManager Server, problems could arise.

The general rules for processing UANodeSet XML file are described in OPC 10000-6 Annex F.14 This clause provides the following instructions related to NodeIds:

“Variants can contain NodeIds, ExpandedNodeIds and QualifiedNames that must be modified, so the NamespaceIndexes and ServerIndexes reference the NamespaceUri and ServerUri tables in the UANodeSet.”

When generating a UANodeSet XML file, the NamespaceArray is included at the top of the UANodeset XML file. The NamespaceIndex may be updated to match the list of Namespaces listed at the top of the generated file. But in this case, not all Namespace entries might be defined in the ConnectionManager Server. So, exporting a UANodeset XML from the Server could result in an error (if the index in a Node does not exist in the Server’s NamespaceArray), or it could result in an incorrect Namespace being referenced. The simplest solution is to ensure that the ConnectionManager Server include all Namespaces in its NamespaceArray, even if the only Node using that Namespace is in the complex data structure.

When an engineering tool generates a ConnectionConfigurationSet, if it includes DataSets, then it will need to ensure that any NodeId it generates for its ConnectionManager’s Server does not have a translation issue. Multiple options exist:

• This could be done by ensuring that the ConnectionManager Server configuration includes all Namespaces used for all NodeIds, even if the NodeIds are not on the ConnectionManager’s Server.
• The ConnectionManager’s Server could understand the ConnectionConfigurationSet PubSubConfiguration and bypass standard processing for the PubSubConfiguration (i.e., not map).
• The engineering tool does not provide DataSets that include Variables, and the ConnectionManager generates the PubSubConfiguration.

The actual manner in which this is handled is vendor-specific, but the developer should be aware of these related issues.