WorkSpace (-W)

When you run a map, the Command Server uses its own paging scheme to access information about the data as efficiently as possible. Portions of both workspace data and actual data are paged as needed. The workspace can be either a combination of files and memory or memory-based. When file-based workspace is used, the amount of memory accessed at a given time depends on the number of pages and the count of pages configured for map execution.

When the WorkSpace execution command (-W) is used without other specified options, a file-based WorkSpace is used and the work files are named using the map name; they are placed in the same directory as the compiled map file. If the -W command is used without a corresponding WorkSpace PageSize execution command (-P), the WorkSpace settings in the compiled map are used.

You can optimize the speed of map processing using the workspace configuration setting. When processing a large amount of input data in a map, use the file-based workspace with default paging. When small amounts of data are processed, smaller page sizes and memory-based workspaces are recommended. When using burst processing, adjust the workspace configuration for the size of burst data.


-W[M]|{[D][U][=dir]}

Option

Description

D

A file-based WorkSpace is used. The work files are deleted after a map is executed. All work files are deleted, except those being saved for reuse as specified on the input cards for each map.

U

A file-based WorkSpace is used. A unique name is generated for each work file created and the files are automatically deleted after the map is executed. Unique work file names are useful if you are running more than one instance of the same map at the same time. Each time a map is run, a work file is created with a unique name, which eliminates any conflict between work files.

Unique workfiles are automatically deleted upon completion. The format of the unique workfile name is:

Mer_mapname_processID_time_xxxxxxx_hostname[I|O].cardnumber

mapname: The executable map name.
processID: The process id from the operating system.
time: The time when the process started, in number of seconds since midnight on January 1, 1970.
xxxxxxx: The mapping instance for the process, as an incremental hex number starting at 1 (thru 0xFFFFFFF).
hostname: The host name of the computer where HCL Link is running.
[I|O]: I is an input and O is an output
cardnumber: The input or output card number.

M

A memory-based WorkSpace is used. This might enhance performance for specific maps depending on work file requirements and the amount of available memory by keeping the work file information in memory and eliminating the I/O necessary with file-based work files. If there is insufficient memory available to build the work file, the map returns a Disk write error message.

When this option is used, none of the other work file options can be used. For example, D - deleting work files, U - unique work file names, and =dir - specifying the directory.

=dir

Specify the directory path in which you want the work files to be created. By default, work files are created in the same directory as the compiled map (.mmc) file.