SYSREC data sets in LOADPLUS

SYSREC is an input data set that contains the input data that you are loading and, if applicable, the name of a file that contains LOB or XML data. LOADPLUS always requires at least one SYSREC data set.

LOADPLUS uses the data set in the PRELOAD phase for two-phase load or in the COMBINED phase for single-phase processing. At the end of the PRELOAD phase or the COMBINED phase, LOADPLUS frees this data set. However, you must specify a SYSREC DD statement in the JCL (which can be SYSREC DD DUMMY) or ensure that your LOAD command contains the INDSN option if you restart the job, regardless of the phase in which you restart.

Related topics

LOADPLUS-DD-statements

Building-the-LOADPLUS-job

You can use either form of the default data set name (SYSREC or SYSREC nn). For information about how to specify DD statements when using multiple data sets, see Using multiple SYSREC data sets. To override the default ddname, use the INDDN command option.

As an alternative to specifying SYSREC data sets in your JCL, you can have LOADPLUS dynamically allocate your SYSREC data sets by using the INDSN command option.

Requirements and restrictions

SYSREC data sets have the following general requirements and restrictions. For more restrictions when using multiple SYSREC data sets, see Using multiple SYSREC data sets.

  • You must use a single SYSREC DD in the following cases:
    • When you specify ORDER PRESORTED
    • If you are generating values into an identity column and you want the values ordered in input order
  • We strongly recommend that you do not use RECFM=FBS.
  • If you are using RECFM=U, or have concatenated an uninitialized data set in your SYSREC DD (which results in RECFM=U), LOADPLUS processes only one record per block and produces unpredictable results.
  • A data set that is in the spanned-record format must be allocated as RECFM=YES with LRECL=32768.
  • LOADPLUS does not support SYSREC data sets allocated as DSNTYPE=LARGE.
  • Do not specify FREE=CLOSE for SYSREC data sets. 

Using multiple SYSREC data sets

LOADPLUS permits up to 256 SYSREC DD statements, providing multitasking of the read process in the PRELOAD phase or in the COMBINED phase. When specifying multiple data sets, you can use one or both forms of the data set name (SYSREC or SYSRECnn).

Important

If LOADPLUS is dynamically allocating your input data sets and you need more than 99 data sets, use the INDDN command option to override the default data set name of SYSREC. Specify a ddname prefix that results in eight characters or less after LOADPLUS appends the highest data set number. For more information, see Specifying ddname prefixes.

LOADPLUS reads the SYSREC data sets in parallel only when loading data into a partitioned table space. For details about performance implications of multitasking in LOADPLUS, see Enabling-multitasking-for-performance-in-LOADPLUS.

If you specify PRELOAD ANALYZE, LOADPLUS stops after the optimization process of the PRELOAD phase and displays a message that indicates the optimal number of sort and read tasks that it can process concurrently. You can use this information to adjust the number of SYSREC data sets in your JCL.

Considerations

The following considerations apply to multiple SYSREC data sets:

  • Data in the SYSREC data sets does not have to be grouped by partition.
  • Data for a particular partition can be in one or more SYSREC data sets.
  • Although not required, we recommend that you order your SYSREC DD statements from the largest data set to the smallest to optimize performance.

Requirements

In addition to the general requirements and restrictions on SYSREC data sets, the following requirements apply to multiple SYSREC data sets:

  • You can specify a mixture of DASD files and pipes as input. However, you cannot concatenate these files within one SYSREC DD.

  • If you concatenate your input data sets, specify the data set with the largest block size and record length first.

    Important

    Alternatively, on the first data set in the concatenation, specify values for the BLKSIZE and LRECL parameters that will ensure that this data set has the largest block size and record length.

  • If you use multiple pipe data sets, you must specify a separate SYSREC DD statement for each pipe data set, and the DCB parameters for all of the pipe data sets must be the same.
  • Multiple SYSREC data sets must use the same record format (RECFM). In other words, they must all be allocated as either F or FB, or they must all be allocated as V, VB, or VBS. Mixing fixed and variable format data sets produces unpredictable results.
  • Multiple SYSREC data sets that are defined with F or FB record format must use the same record length (LRECL).

Specifying files that contain LOB or XML data

In your SYSREC file, you can specify the name of a PDS, PDSE, or HFS file that contains your LOB or XML data. Note the following requirements:

  • If your referenced file is a PDS or PDSE, you must include a member name in the name provided in the input file.

  • If your referenced file is an HFS file, your input file must specify the full path name of the file, as in the following example:

    /home/rdajdm/lobfr1


 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*