HPC_Partition method
Collects the information that you need to initiate a single calculation step for an Excel workbook on an HPC cluster.
Syntax
Function HPC_Partition() As any base type or Variant[]
Parameters
This method has no parameters.
Return value
Specifies the information that is needed to perform a single calculation step in the Excel workbook. For example, if you want to split up the calculation by rows, then this return value could be the number of the row to use for the calculation.
When the last calculation step is complete, this macro should return Null.
The type of this return value must match the type of the parameter that your implementation of the HPC_Execute macro takes.
Remarks
You implement this macro for an Excel workbook for which you want to run calculations on an HPC cluster as part of the partition, calculate, and merge macro framework. This macro represents the partition stage of that framework. The framework implements an iterative calculation model for the workbook, where different calculation steps are performed on different nodes in the HPC cluster. The partition stage of the framework establishes how the overall calculation is split into individual calculation steps.
The HPC Services for Excel client library calls this macro multiple times for a workbook calculation, once for each calculation step.
This macro runs on the client computer. When the return value of this macro is sent to the HPC_Execute macro, that information is sent through the HPC Job Scheduler Service from the client computer to the compute node that performs the calculation step. This data transmission occurs asynchronously, so that HPC_Execute does not necessarily complete one calculation step before the next iteration of HPC_Partition sends the information for the subsequent step.
To improve performance, you should minimize the work that you perform in the HPC_Partition macro, including the calculation and display tasks. For example, you should open external resources such as log files and database connections in the HPC_Initialize macro rather than in the HPC_Partition macro. You should also avoid reading large sets of data in the HPC_Partition macro.
Examples
For example implementations of the HPC_Partition macro, see the following resources:
Accelerating Excel 2010 with Windows HPC Server 2008 R2: Building VBA applications and workbooks for a Windows HPC Cluster (https://go.microsoft.com/fwlink/p/?linkid=205876)
The macros in the VBASample.xlsm workbook that is included in the Excel\ExcelRunner folder in the sample code package for the Windows HPC Pack 2008 R2 SDK (https://go.microsoft.com/fwlink/p/?linkid=205875)
Requirements
Product |
HPC Pack 2008 R2 Client Utilities |