Each data catalog entry has a home server—the grid server where the information about that object is stored. But the data catalog itself, and all the services and objects that it contains, are distributed among the various servers that comprise your Data Federation domain.
The home server is the grid server on which the information that Data Federation maintains about that object resides—its ACL, attributes, etc.
When you first create your Data Federation domain, a number of core services and objects are created on the grid domain controller (GDC) and registered in the data catalog. When you join a new grid server to a Data Federation domain, a number of core services and data catalog directories are created on that new server, and the top level data catalog directories on that server are linked into the appropriate places on the GDC.
In some cases, such as when you create a database connector or a data service, Data Federation gives you the option of choosing the home grid server for that object. In other cases, the choice is implicit—for example, when you create a database operation, its home server is the same as the home server of the database connector with which it is associated.
Thus, the data catalog as well as all the services and objects that it contains are distributed, not replicated, among the various servers that comprise your Data Federation domain.
When you create a data catalog directory, the directory is created on the same grid server as the directory that contains it—so the taxonomies you create will all reside on the GDC, since the root of the data catalog is on the GDC.
Send your feedback on this help topic to Sybase Technical Publications: pubs@sybase.com
Your comments will be sent to the technical publications staff at Sybase, Inc. For product-related issues or technical support, contact Sybase Technical Support at 1-800-8SYBASE.