SORTWORK
The SORTWORK keyword is valid with the BUILD, CLONE, RELOAD, REORG (Online Reorg function and Reorg function), and GROUP commands and with the HD Sort utility.
You can specify the number of SORTWORKnn data sets, the number of primary cylinders, and the unit name for this option in the environment setup member (DLIGSET0). You can specify all parameters in PLUSIN. The following table describes the keyword.
To do this | Specify this in PLUSIN |
|---|---|
Dynamically allocate sort work data sets. | SORTWORK(nn,cccc,unitName,dataclas,mgmtclas,storclas,alloctype,secalloc,sortpref) |
Internal default: (3,100,SYSDA) | |
The SORTWORK keyword enables you to set parameters for dynamically allocated sort work data sets. If you do not specify the SORTWORK keyword, the default is used. You can specify the SORTWORK keyword as follows:
The SORTWORK parameters are positional, and they must be separated by commas.
Parameter | Description |
|---|---|
nn | Number of SORTWKnn data sets (0 through 99). The default value is 3. If you specify 0, no data sets are allocated. |
cccc | Number of primary cylinders to allocate for each data set (1 through 32767). The default value is 100. |
unitName | Unit name for sort work data sets. Specify any valid unit name. The default value is determined by what you specified in the environment setup member. |
dataclas | SMS data class construct. Specify any valid data class name. The default value is determined by what you specified in the environment setup member. |
mgmtclas | SMS management class construct. Specify any valid management class name. The default value is determined by what you specified in the environment setup member. |
storclas | SMS storage class construct. Specify any valid management class name. The default value is determined by what you specified in the environment setup member. |
alloctype | Allocation type. The acceptable values are C, CYLS, T, and TRKS. |
secalloc | Number of secondary cylinders to allocate for each data set (numeric from 1 to 9999). |
sortpref | Exactly two characters, a unique prefix for any SORTWORK DDs to be used by this command. |
The SORTWORK keyword applies to each sort.
For the secondary index sorts, the utility dynamically allocates the sort work data sets ddnames in the following format:
The variablex is the sort task that is running (1, 2, or 3) andnn is the number of sort work data sets.
If you specify SORTWORK(0, x,x) and the amount of data to be sorted is too large for an incore sort, you must include SRTxWKnn DD statements in the JCL.
If you specify PARALLEL(YES), the Index Build function chooses the optimum number of sorts (as many as three).
Related topic