Accessing Local and Remote Data

You can access data using the following methods:

You can access data implicitly in the context of a job. When code is generated for a job, it is generated in the current context. The context includes the default SAS Application Server when the code was generated, the credentials of the person who generated the code, and other information. The context of a job affects the way that data is accessed when the job is executed.

In order to access data in the context of a job, you need to understand the distinction between local data and remote data. Local data is addressable by the SAS Application Server when code is generated for the job. Remote data is not addressable by the SAS Application Server when code is generated for the job.

For example, the following data is considered local in the context of a job:

The following data is considered remote in a SAS Data Integration Studio job:

Remote data has to be moved because it is not addressable by the relevant components in the default SAS Application Server at the time that the code was generated. SAS Data Integration Studio uses SAS/CONNECT and the UPLOAD and DOWNLOAD procedures to move data. Accordingly, it can take longer to access remote data than local data, especially for large data sets. It is especially important to understand where the data is located when using advanced techniques such as parallel processing because the UPLOAD and DOWNLOAD procedures run in each iteration of the parallel process.

For information about accessing remote data in the context of a job, administrators should see the section on "Multi-Tier Environments" in the "SAS Data Integration Studio" chapter of the SAS Intelligence Platform: Desktop Application Administration Guide. Administrators should also see Using Deploy for Scheduling to Execute Jobs on a Remote Host. For details about the code that is generated for local and remote jobs, see the subheadings about LIBNAME statements and remote connection statements in Common Code Generated for a Job.

When you use SAS Data Integration Studio to access information interactively, the server that is used to access the resource must be able to resolve the physical path to the resource. The path can be a local path or a remote path, but the relevant server must be able to resolve the path. The relevant server is the default SAS Application Server, a server that has been selected, or a server that is specified in the metadata for the resource.

For example, in the external file wizards, the Server tab in the Advanced File Location Settings window enables you to specify the SAS Application Server that is used to access the external file. This server must be able to resolve the physical path that you specify for the external file.

As another example, assume that you use the Open option to view the contents of a table in the Inventory tree. If you want to display the contents of the table, the default SAS Application Server or a SAS Application Server that is specified in the library metadata for the table must be able to resolve the path to the table.

In order for the relevant server to resolve the path to a table in a SAS library, one of the following conditions must be met:

You can use the Data Transfer transformation to move data directly from one machine to another. Direct data transfer is more efficient than the default transfer mechanism.

For example, assume that you have the following items:

You can use SAS Data Integration Studio to create a process flow diagram that moves data from the source on machine 1 to the target on machine 3. By default, SAS Data Integration Studio generates code that moves the source data from machine 1 to machine 2 and then moves the data from machine 2 to machine 3. This is an implicit data transfer. For large amounts of data, this might not be the most efficient way to transfer data.

You can add a Data Transfer transformation to the process flow diagram to improve a job's efficiency. The transformation enables SAS Data Integration Studio to generate code that migrates data directly from the source machine to the target machine. You can also use the Data Transfer transformation with a SAS table or a DBMS table whose table and column names follow the standard rules for SAS names.