Chapter Contents


SAS/CONNECT User's Guide

Benefits of Data Transfer Services

A major benefit of data transfer services is to offload work from a remote system to one or more local machines and boost response time for production systems that run in the remote environment. After the data is downloaded to the local machine, all subsequent data access and processing is done by the local processor.

By moving a copy or subset of the original data to the local machine, production applications can be run on the local machine without further remote CPU consumption or impact on remote production data systems.

Moving a copy of the data to your local system adds robustness to your decision support environment. In the case of a network failure that would temporarily eliminate access to the remote data, you can continue working with your local copy of the data.

You can transfer only the data that you need by using WHERE processing, data set options, or both to dynamically subset the data as it is being transferred to the local machine. This reduces network traffic and gives you only the data that you need on the local system.

Data transfer services not only facilitate moving data from a larger source to a requesting local workstation, but also support the model of a centralized control point, such as a mainframe, which initiates communication to a network of workstations.

This model enables centralized distribution of data and applications. Automated jobs can be run during non-peak hours to distribute data and applications to multiple machines that need the data and the applications for the next day's work. Similarly, jobs can be set up to query a network of workstations for the purpose of gathering data and storing it in a centralized repository.

Another reason to use data transfer services is for backup purposes. Data and applications can be copied from a local system that has small memory resources to a remote system that has more memory resources. This provides a backup in case of loss on the local system.

Applications developers can also make use of data transfer services. Many applications are developed to run on a remote host and take advantage of its compute services. Data transfer services reduce the workload on your remote host during development of these applications because you can use the local host as a program development tool. You can perform program editing, testing, and debugging by using the local host's resources instead of the remote host's resources.

The development environments at many computing installations often have a higher number of users who work on one system than on other systems. On the system with the heaviest load, response time, execution queues, and other performance factors are less efficient because so many people are running applications concurrently.

Using data transfer services, you avoid contention for heavily used host resources by creating and testing SAS programs on a less busy system (your local system), and then remote-submitting the fully developed and tested program to the heavily loaded system (your remote system).

Each time you execute a program in a local host SAS session for testing purposes, you avoid adding to the load on the remote host computer. This method can result in significant savings of remote host resources and convenience for you.

For example, suppose you are developing a SAS program that will run as a production program on the remote host. Your program analyzes data in a SAS data set on the remote host system and creates several reports from the analysis information. To run many tests of the program before it is perfected and avoid the delays involved in working on the remote host, create and store the SAS program on your local host. Test the program by downloading the SAS data set that is analyzed by the program, or test the program by using data that is stored in the local host. After the program is complete and correct, upload the program file to the remote host.

Chapter Contents



Top of Page

Copyright 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.