Available with Standard or Advanced license.
There are important considerations to be aware of when replicating datasets that participate in advanced geodatabase functionality. These considerations are described in the following sections.
Attribute rules and contingent values
When you replicate feature classes and tables with attribute rules or contingent values applied, the feature classes and tables are replicated to the child geodatabase; however, the attribute rules and contingent values are not copied to the child geodatabase and only persist on the data in the parent geodatabase.
When synchronizing replicas from the child geodatabase to the parent geodatabase, if edits made to the data in the child geodatabase violate an attribute rule or contingent value in the parent geodatabase, an error will occur, indicating that synchronization failed. If an error occurs during synchronization, the operation is rolled back, and a message indicating the synchronization failure is recorded in the replica log. Any changes previously applied are removed, and the system returns to its state before the changes were synchronized.
Geometric networks are read-only and are not supported in a replica created in ArcGIS Pro; this includes any feature classes that participate within the geometric network.
If your data contains representations before replica creation, they will be included in the replica. If you apply a symbol to a feature in one replica and synchronize, the symbol will be applied to the feature in the relative replica. Modifications to representations are not supported.
For example, consider the case in which a new symbol is added to a representation and applied to a feature in a replica. When that feature is synchronized, it draws with an unknown symbol in the relative replica, because the new symbol does not exist in the relative replica. It is also possible to add a different symbol with the same ID to the relative replica. In this case, the synchronized feature draws with the incorrect symbol in the relative replica.
Terrains and network datasets
When replicating network datasets and terrains, only the feature classes used to build these data structures can be replicated. The actual terrain data structures, terrain, and network datasets are not replicated. For feature classes that participate in a network dataset or terrain to be replicated, they must be registered as versioned and meet data requirements for replication. You can rebuild the terrain or network dataset in the child replica's geodatabase after replica creation.
Regarding network datasets, turn feature classes and traffic and time zone tables are not replicated. However, signpost feature classes are replicated.
When replicating topology datasets, all feature classes that participate in a topology are replicated together. This means that individual topology feature classes cannot be excluded from a replica.
When topology feature classes are replicated, the whole extent of the topology in the child replica is marked as dirty. To discover existing errors, the topology must first be validated. The topology in the child replica behaves in the same way as in the parent replica; edits create dirty areas and validation creates and deletes errors.
If a feature is marked as an exception to the topology rules, this is maintained and will appear as an exception in the child replica.
The following example shows the behavior of replicating a topology. When the topology is first replicated, the entire extent is marked as dirty on the child replica. After validation, the original error feature is rediscovered. The feature marked as an exception to the topology rules maintains its exception status on the child replica after the replication process.