What Business Problems Can a Grid Solve?

An organization might have multiple users submitting jobs to run on one server. When the environment is first configured, the server might have been sufficient to handle the number of users and jobs. However, as the number of users submitting jobs grows, the load on the server grows. The increased load might lead to slower processing times and system crashes. In a SAS grid environment, jobs are automatically routed to any one of the servers on the grid. This spreads the computing load over multiple servers, and diminishes the chances of a server becoming overloaded. If the number of jobs exceeds the resources available, the jobs are queued until resources become available. If the number of users continues to increase, you can increase capacity by adding servers to the grid.

Your organization might have services and long-running SAS programs that are critical to your operations. The services must be available at all times, even if the servers that are running them become unavailable. The SAS programs must complete in a timely manner, even if something happens to cause them to fail. For a SAS program that takes a long time to run, this means that the program cannot be required to restart from the beginning if it ends prematurely.

You can configure the critical services within your SAS grid environment to be highly available. SAS Grid Manager can monitor the critical services, detect if they fail or if the machine on which they are running fails, and automatically start the services on a failover host. Either a hardware load balancer or DNS name resolution is used to redirect clients to the service running on the failover host. This ensures that critical services remain available to clients without any manual intervention.

By using options on the SAS Grid Manager Client Utility, you can specify that SAS programs submitted to the grid are automatically restarted from the point where they stopped if they end before completion. The job restarts from the last completed procedure, DATA step, or labeled section. Jobs that take a long time to run do not have to start over at the beginning. You can also use the restart capability with queue options that automatically requeue jobs that end prematurely to provide a complete high-availability environment for SAS programs.

Your organization might have a process running to analyze a certain volume of data. Although the server that is processing the job is sufficient to handle the current volume of data, the situation might change if the volume of data increases. As the amount of data increases, the load on the server increases, which can lead to longer processing times or other problems. Changing to a larger-capacity server can involve considerable expense and service interruption.

A SAS grid environment can grow to meet increases in the amount of data processed. If the volume of data exceeds the capacity of a server on the grid, the processing load can be shared by other grid servers. If the volume continues to increase, you can add servers to the grid without having to make configuration changes to your processes. Adding servers to the grid is also more cost-effective than replacing a single large server, because you can add smaller servers to handle incremental increases in data volume.

Your organization might have a process running to perform a certain level of analysis on data. If you want to increase the complexity of the analysis being performed, the increased workload puts a greater strain on the processing server. Changing the computing power of the server involves considerable expense and interrupts network availability.

Using a SAS grid environment enables you to add computing power by adding additional computers on the grid. The analysis job can be divided up among the grid nodes, which enables you to perform more complex analysis without increasing the load on any single machine.

Your organization's ability to perform the data analysis that you need depends on a flexible computing infrastructure. You must be able to add needed resources quickly and in a cost-effective manner as the load increases. You must also be able to handle maintenance issues (such as adding or replacing resources) without disrupting your work. A SAS grid environment enables you to maintain a flexible infrastructure without disrupting your operations.

As your data-processing needs grow, you can incrementally add computing resources to your grid by adding smaller, less-expensive servers as new server nodes. This ability prevents you from having to make large additions to your environment by adding large and expensive servers.

When you need to perform maintenance on machines in the grid, the grid can still operate without disruption. When you take the servers offline for maintenance or upgrades, SAS Grid Manager routes to work to the machines that are still online. Users who send work to the grid for processing do not have to change their way of working. Work that is sent to the grid is processed just as before.

Likewise, the SAS grid environment adapts if a computer fails on the grid. Because SAS Grid Manager automatically avoids sending work to the failed machine, the rest of the grid is still available for processing and users do not see any disruption.