Parallel Processing Parameters

Payroll processes are designed to take advantage of multiprocessor computers. You can improve performance of your batch processes, such as Calculate Payroll or Calculate Gross Earnings, by splitting the processing into a number of threads, or subprocesses, which run in parallel.

To improve performance you can also set the number of payroll relationship actions that process together and the size of each commit unit for the batch process.

Here’s a list of parameters that you can use for parallel processing.

Parallel Processing Parameters

Threads

When you submit a batch process, the Threads parameter determines the total number of subprocesses that run concurrently. The number of subprocesses equals the Threads value minus 1.

Set this parameter to the value that provides optimal performance on your computer:

• The default value of 1 is set for a single-processor computer.

• Benchmark tests on multiprocessor computers show that the optimal value is approximately 2 processes per processor.

  For example, if the server has six processors, set the initial value to 12 and test the impact on performance of variations on this value.

Chunk Size

The Chunk Size parameter:

• Indicates the size of each commit unit for the batch process.

• Determines the number of assignment actions that are inserted during the initial phase of processing.

• Sets the number of assignment actions that are processed at one time during the main processing phase.

This parameter doesn't apply to all processes, such as Generate Check Payments and Retroactive Pay.

To set the value of the Chunk Size parameter, consider the following points:

• Parameter values range from 1 to 16,000.

• The default value is 20, which was set as a result of benchmark tests.

• Each thread processes one chunk at a time.

• Large chunk size values aren't desirable.