SAS Grid Computing Basics

A SAS grid computing environment is one in which SAS computing tasks are distributed among multiple computers on a network, all under the control of SAS Grid Manager. In this environment, workloads are distributed across a grid of computers. This workload distribution enables the following functionality:
Workload balancing
enabling multiple users in a SAS environment to distribute workloads to a shared pool of resources.
Accelerated processing
allowing users to distribute subtasks of individual SAS jobs to a shared pool of resources. The grid enables the subtasks to run in parallel on different parts of the grid, which completes the job much faster.
Scheduling jobs
allowing users to schedule jobs, which are automatically routed to the shared resource pool at an appropriate time.
SAS Grid Manager provides load balancing, policy enforcement, efficient resource allocation, prioritization, and a highly available analytic environment for SAS products and solutions running in a shared grid environment. It also separates the SAS applications from the infrastructure used to execute the applications. This enables you to transparently add or remove hardware resources as needed and also provides tolerance of hardware failures within the grid infrastructure. SAS Grid Manager integrates the resource management and scheduling capabilities of the Platform Suite for SAS with the SAS 4GL syntax and subsequently with several SAS products and solutions.
SAS Grid Manager includes these components, as illustrated in Grid Topology. :
Grid Manager plug-in
a plug-in for SAS Management Console that provides a monitoring and management interface for the jobs and resources in your grid
grid syntax
the SAS syntax necessary to grid-enable the SAS workload
Platform Suite for SAS
components provided by Platform Computing to provide efficient resource allocation, policy management, and load balancing of SAS workload requests.
The Platform Suite for SAS includes these components:
Load Sharing Facility (LSF)
this facility dispatches all jobs submitted to it, either by Process Manager or directly by SAS, and returns the status of each job. LSF also manages any resource requirements and performs load balancing across machines in a grid environment.
Process Manager (PM)
this is the interface used by the SAS scheduling framework to control the submission of scheduled jobs to LSF and manage any dependencies between the jobs. Process Manager includes two optional components, Calendar Editor and Flow Manager.
Calendar Editor is a scheduling client for a Process Manager server. It enables you to create new calendar entries for time dependencies.
Flow Manager provides a visual representation of flows that are created and scheduled through the Schedule Manager plug-in as well as reports scheduled through SAS Web Report Studio. Flow Manager enables you to view and update the status of jobs in a flow and rerun jobs.
Grid Management Services (GMS)
this is the interface to the Grid Manager plug-in in SAS Management Console. It provides the run-time information about jobs, hosts, and queues for display in SAS Management Console.
Platform MPI
this is a high-performance implementation of the Message Passing Interface (MPI) standard for both the Linux and Microsoft Windows operating systems. It provides the middleware used by grid-enabled SAS procedures.
Platform RTM for SAS
a Web-based tool that enables you to graphically view the status of devices and services in a SAS grid environment as well as manage the policies and configuration of the grid. This application is not part of Platform Suite for SAS, but can be downloaded separately from http://www.sas.com/apps/demosdownloads/platformRTM_PROD__sysdep.jsp?packageID=000669