Using RLS to Access Types of Data

RLS supports access to the following types of data:

*Catalog update is not supported if the computers that the client and the server run on do not have compatible architectures.

In order for a client to use RLS to update a catalog on a server, the architectures of the computers on which the client and the server run must be compatible. If computer architectures are incompatible, the following error message is displayed:

ERROR: You cannot open catalog name through 
       server ID because write access to 
       catalogs is not supported when the user 
       machine and server machine have different 
       data representations.

RLS and a SAS/CONNECT single-user server support update access to data that is stored in an external database. The SAS/ACCESS engines and the SQL engine recognize the single-user server as one user and, therefore, enable update access for external database sources.

However, SAS/ACCESS engines and the SQL engines prohibit update access to external database sources when using RLS and a multi-user server. Updating is prohibited because of the inability of a multi-user server or a database to detect and manage conflicting requests from multiple users. A detection facility is necessary in order to generate audit trails and to guarantee data integrity and security.

RLS supports access to SAS views, which include DATA step views, SAS/ACCESS views, and PROC SQL views.

When the server accesses the library that contains the SAS view, the view is interpreted at the server by default. The server loads and calls the engine that is appropriate to the SAS view to read and transform the underlying data. The processing that is required to generate the SAS view is performed at the server, and the resulting SAS view is transferred to the client with a minimum cost to the network. Client resources are not used to interpret the SAS view.

For all PROC SQL views or for any other type of SAS view that is processed between a client and a server whose computer architectures are compatible, the SAS view can be interpreted at the client. To interpret a SAS view at the client instead of at the server, set the RMTVIEW= option to NO in a LIBNAME statement. Here is an example:

libname payroll rmtview=no server=wntnode;

For DATA step views and SAS/ACCESS views, if the architectures of the computers that the client and the server run on are different, the views can be interpreted only at the server.

In order for a client to use RLS to access a SAS utility file of the type PROGRAM or ACCESS on a server, the architectures of the computers that the client and the server run on must be compatible. If computer architectures are incompatible, the following error message is displayed:

ERROR: You cannot open utility file name through 
       server ID, because access to utility 
       files is not supported when the user machine 
       and server machine have different data 
       representations.

A SAS utility file of the type PROGRAM contains compiled DATA step code, which cannot be processed at the client. The DATA step can be executed at the server if the DATA step is referenced by a DATA step view that is interpreted at the server.