Overview of Workspace Servers and Stored Process Servers

Stored process servers and the workspace servers are crucial elements of the SAS Intelligence Platform that enable clients to perform SAS processing and to access Foundation SAS resources.

SAS Stored Process Servers interact with SAS by submitting stored processes, which are SAS programs that are stored and can be submitted by SAS client applications. You can use stored processes to perform complex tasks such as analyzing data and creating reports, and then returning the results to the client or publishing the results to a channel or repository.

Each stored process server process handles multiple users, and by default each server uses multiple server processes or instances. A load-balancing algorithm distributes client requests between the server processes. For more information about load balancing, see Overview of Load Balancing.

If the job load for your stored process server is high, you might want to add additional server processes to your server definition. Each server process is defined as a MultiBridge connection in SAS Management Console.

SAS Workspace Servers interact with SAS by creating a server process for each client connection. The workspace server process is owned by the client user who made the server request. Each workspace server process enables client programs to access SAS libraries, perform tasks by using the SAS language, and retrieve the results.

In the default configuration, SAS presents you with a workspace server that is not pooled (standard workspace server) and a workspace server that is pooled (pooled workspace server).

For the standard, non-pooled workspace server, SAS creates a new server process each time that a client requests a connection. For simple configurations and for sites that do not place heavy loads on a workspace server, a non-pooled server might be adequate. In situations where the demands on the workspace server are greater, you should (depending on which SAS application you are using) consider configuring the workspace server for either pooling (Web applications) or load balancing (desktop applications).

In a pooling configuration, a set of server processes are reused to avoid the processing time that is associated with starting a new process for each connection. SAS offers two types of pooling: client-based and server-based. A pooling configuration can also be shared across multiple machines. Pooling is recommended if your server supports SAS Web Report Studio and other Web applications. For more information, see Overview of Pooling.

In a load-balancing configuration, your workspace server processes are distributed between multiple machines. Load balancing is recommended if your server supports applications that submit large jobs, such as SAS Data Integration Studio. For more information, see Overview of Load Balancing.

SAS Pooled Workspace Servers are workspace servers in every respect except that these servers automatically use pooling and load balancing. Like a standard workspace server, each pooled workspace server enables client programs to access SAS libraries, perform tasks by using the SAS language, and retrieve the results. For more information, see How Server-side Pooling Works.

When the installer at your site runs the SAS Deployment Wizard, that person defines metadata for a SAS Application Server.

Usually, your application server contains a stored process server, a workspace server, and a pooled workspace server.

The initial stored process server is configured as a load-balancing server named by default SASApp - Stored Process Server. By default, the stored process server definition includes three MultiBridge connections.

The initial workspace server is configured as a standard workspace server named SASApp - Workspace Server.

The initial pooled workspace server is configured as a load-balanced, server-based pooled workspace server named by default SASApp - Pooled Workspace Server.