Hot Topics

Introduction to SAS Grid Computing

Demands on IT to deliver a scalable, reliable environment and faster response times continue to increase as more data, users and applications are being used to solve complex business problems. Enterprise SAS Business Analytic deployments drive the need for:

SAS Grid Manager provides a shared, centrally managed analytic computing environment that is high availability and accelerates processing. It provides workload management to optimally process multiple applications and workloads to maximize overall throughput. A SAS Grid environment also provides the flexibility to incrementally grow the computing infrastructure as the number of users and the size of data increase over time as well as the ability to do rolling maintenance and upgrades without any disruption to the user community. SAS products and solutions that are deployed with SAS Grid Manager are able to leverage the following functionality in a distributed grid environment:

Enterprise scheduling of production jobs is an important and necessary function in every enterprise. SAS Grid Manager includes the Schedule Manager plug-in with SASŪ Management Console so you can create SAS workflows and schedule them based on time and file events. Multiple workflows as well as the jobs within a single workflow are then distributed to the SAS grid environment. SAS jobs can be created by a variety of SAS applications and solutions or written by SAS programmers. Various levels of scheduling capabilities have been incorporated directly into many SAS products and solutions, including SAS Data Integration Studio, SASŪ Marketing Automation, SASŪ Marketing Optimization, and SASŪ Web Report Studio. The jobs created by SAS products and user-written SAS programs can be used to create simple or complex workflows and Schedule Manager can be used to schedule them to a SAS grid environment.

Workload management is critical when you have multiple users running multiple SAS applications which is typically the case for most SAS enterprise analytic environments. SAS Grid Manager provides a centrally administered environment to enable policies that determine which users and/or jobs get priority, which jobs get which share of the computing resources, and prevent a mass submission of work requests resulting in extremely poor performance or even frequent server crashes. SAS Grid Manager maps the SAS work requests to the available resources. If necessary, it runs a subset of the work and queues the remaining work for execution as soon as resources become available. High-priority jobs can even preempt low-priority work so that the most critical business processes execute first. SAS Grid Manager provides the structure to create an organized and managed SAS analytic environment. This ensures that the appropriate resources are allocated to the appropriate users or appropriate workload to meet the objectives of the entire organization.

Parallel workload management is important for the subset of SAS applications that consist of sub-tasks that are independent units of work and can be distributed across a grid and executed in parallel. The primary benefit of distributed parallel execution is substantial acceleration of the entire application. Several SAS products, such as SAS Data Integration Studio and SAS Enterprise Miner, have code generation engines that will recognize opportunities for parallelization and generate the appropriate code to submit the flow to SAS Grid Manager to execute in parallel across the grid. In addition, PROC SCAPROC can be used to run existing SAS programs and generate a new version of the program that is grid-enabled and contains the appropriate syntax to execute segments of the program, when possible, in parallel across the grid.

For more information about grid computing implementations, please see the SAS Grid FAQ document.

page divider

We at SAS have created the Scalability Community to make you aware of the connectivity and scalability features and enhancements that you can leverage for your SAS installation. The success of this community depends on you. Send electronic mail to with your comments, requirements, and suggestions.