REORG (Parallel BACKUP/LOAD) › How to Use

How to Use

The MUF must be active when you execute this command. Execute the REORG function using the following command format.

BACKUPONLY REORG

►►─ REORG OPTION2=BACKUPONLY,DBID=n,AREA=a,DDNAME=d,INNUM=n ──────────────────►

 ►─┬────────────┬─┬──────────────┬────────────────────────────────────────────►◄
   └─ ,CYCLE=n ─┘ └─ ,SEQBUFS=n ─┘

BACKUPLOAD/LOADONLY REORG

►►─ REORG OPTION2= ─┬─ BACKUPLOAD ─┬─ ,DBID=n,AREA=a,DDNAME=d,INNUM=n ────────►
                    └─ LOADONLY ───┘

 ►─┬─────────────────────────────────────┬─┬─────────────┬────────────────────►
   └─ ,FORMAT= ─┬─ BACKUP ◄ ───────────┬─┘ └─ ,KBYTES=n ─┘
                └─ EXTRACT,TABLE=name ─┘

 ►─┬────────────┬─┬─────────────┬─┬──────────────────────────┬────────────────►
   └─ ,SLACK=n ─┘ └─ ,NUMBER=n ─┘ └─ ,RESETRCE= ─┬─ YES ◄ ─┬─┘
                                                 └─ NO ────┘

 ►─┬────────────┬─┬────────────┬─┬────────────────────────┬───────────────────►◄
   └─ ,CYCLE=n ─┘ └─ ,TASKS=n ─┘ └─ ,MULTUSE= ─┬─ NO ◄ ─┬─┘
                                               └─ YES ──┘

Note: FORMAT=EXTRACT is only valid when OPTION2=LOADONLY.

Command

REORG

Invokes the REORG function.

Required Keywords

OPTION2=

Specifies the type of REORG.

Valid Entries:

BACKUPLOAD, BACKUPONLY, LOADONLY

Default Value:

(No default)

,DBID=

Identifies the database selected for REORG.

Valid Entries:

DATACOM-ID of the database

Default Value:

(No default)

,AREA=

Identifies the area to be backed up and/or loaded. up and/or loaded.

Valid Entries:

DATACOM-NAME of the area in the database specified in the DBID= parameter

Default Value:

(No default)

,DDNAME=

Specifies the JCL DDname to be used for the data input.

A DDNAME is not acceptable for sequential input or output files if it is a name reserved for a CA Datacom area. Names with the following patterns are therefore not acceptable for DDNAME=:

• 3-byte names that end with XX, meaning they are reserved as either current or future CA Datacom control areas.
• 6-byte names that end with what could be a database ID from 001 through 999.
• 7-byte names that end with what could be a database ID from 1000 through 9999.

  The DDNAME= value is verified for acceptability to protect you from unintentionally causing data corruption. The DDNAME check is the default but optional. You can prevent the DDNAME check by using a DBSIDPR parameter (DBUTLTY_EDIT_DATA_SET=) for individual MUF environments. However, we recommend that you allow the DDNAME check.

  The data corruption risk involves not the DDNAME itself but the content of the data set. For example, suppose that you used the CXX DDNAME as the output of a backup. You then copied the CXX DD statement and changed the DDNAME of the copy to be acceptable, avoiding the DDNAME= error. The backup would, however, then overlay the CXX data set, which is not the intent of a backup.

  If you specify an unacceptable name for DDNAME=, message DB10059E is generated. For more information about DB10059, see the CA Datacom/DB Message Reference Guide.

  Note: We recommend that you allow DDNAME= check protection. You can, however, disable DDNAME= protection. To disable protection, assemble the DBSIDPR module used for this CA Datacom environment and specify NONE for the DBUTLTY_EDIT_DATA_SET= parameter. The default is DBUTLTY_EDIT_DATA_SET=FULL_1, which allows DDNAME= protection. For more information about DBSIDPR and DBUTLTY_EDIT_DATA_SET=, see the CA Datacom/DB Database and System Administrator Guide.

  A reporting only option is available, if you use REORG OPTION2=BACKUPONLY and either DDNAME=* or DDNAME** as described in the following:

  DDNAME=*

  Specifies that a report is to be generated, with the key values selected for each Native Key sequence ID, by providing starting and ending key values, in character and hexadecimal formats. If the starting and ending values are identical, no rows are found. The generated report can be used to assist in key selection if you prefer to back up the data using the FIRSTKEY/LASTKEY options of the BACKUP or EXTRACT functions.

  DDNAME=**

  Specifies a report that can be used for debugging. It shows the full starting key to be assigned to each task, when running the REORG with a proper DDNAME.

  Valid Entries:

  Any name valid to JCL specifications (1-6 characters, the number of the subtask based on INNUM= being suffixed as 1 through 25)

  Note: There is also the special case of the use of * or ** (as previously described).

  Default Value:

  (No default)

,INNUM=

Specifies the number of subtasks to backup and/or load data.

Note: If you are partitioning by some portion of Native Key, running a REORG DDNAME=*,INNUM=n (where n is the number of partitions) gives the current way to break up the data set.

Valid Entries:

A number, 1 through 25

Default Value:

(No default)

Optional Keywords

,CYCLE=

Specifies the minimum number of high-level index entries per task, to divide work. This specification is used during the backup phase of the REORG function.

Valid Entries:

2 through 99

Default Value:

5

,SEQBUFS=

Specifies the number of sequential buffers for each MUF task that is backing up data. Even though allowed by the DBURTBL macro, in this case the value of SEQBUFS= may not be a minus number. To specify minus numbers, add 128. The default of 256 is equivalent to -128 (a negative 128). If you specify less than 8 for SEQBUFS=, the specification is rounded up to a value of 8.

Valid Entries:

0 through 256

Default Value:

256

,MULTUSE=

(For area level DBUTLTY control only.) MULTUSE=YES specifies area level DBUTLTY control that allows a single area in a database to be reorganized with no disabling of the other areas in the database. When MULTUSE=YES:

• The data spaces containing index information are defined with a global scope, making them available to the MUF.
• The index phase must occur in MUF with full interaction with other tasks processing the index and sharing the same buffer pool. Each task does single block reads and writes.
• The area being reorganized cannot have other users for REORG to start and, once started, it cannot have other users to open.
• The MUF needs to be executing on the same image, that is, it cannot be accessed through XCF.

For more information on area level DBUTLTY control, see Area Level DBUTLTY Control.

Valid Entries:

YES or NO

Default Value:

NO

Note: The REORG MULTUSE=YES function requires that the DBUTLTY job run APF-authorized.

,FORMAT=

Specifies the load format, BACKUP or EXTRACT. The OPTION2= keyword must be specified as LOADONLY for FORMAT=EXTRACT to be valid.

Valid Entries:

BACKUP or EXTRACT

Default Value:

BACKUP

,KBYTES=

Specifies the number of 1024-byte blocks of memory each subtask acquires in order to write data rows. We recommend that you specify KBYTES=2000.

Valid Entries:

1 through 9999 (integers only)

This number, multiplied by 1024, is the number of bytes requested through GETMAIN.

Default Value:

four tracks of memory buffering

Note: If a value is specified which results in less than 4 tracks of data, the input is rounded up during execution to be 4 tracks.

,NUMBER=

Specifies the number of contiguous tracks to consider as a group (or stripe). A group of tracks that is acquired for its use, by a subtask, are not available to any other subtask. This requires that the data area must have sufficient free space at completion of the load of the data to allow a partial group for every subtask. You should take this under consideration when when selecting the number of subtasks. Be aware that it is also required that at least one track be available for processing, not in a subtask group, once all the subtasks have completed.

Valid Entries:

Any desired number of tracks

Default Value:

15

,RESETRCE=

Specifies whether you want to keep the TSN (Transaction Sequence Number) that last updated this row. Specify NO to set the TSN to zero. Resetting the TSN to zero is only useful if the number is reset to a lower number during an initialization of the Log Area (LXX), or if the number wraps from FFFFFFFF to zero.

Valid Entries:

YES or NO

Default Value:

YES

,SLACK=

Specifies the number of bytes to reserve, during a load, in each data block, for compressed records to grow. The number entered applies only to blocks loaded to an area during this execution of the load. Blocks not loaded with data use the SLACK value specified in the AREA definition in CA Datacom Datadictionary.

Valid Entries:

A number that is less than the block size

Default Value:

0

,TABLE=

Specifies a table name to indicates that a single table is to be loaded. This parameter is required when FORMAT=EXTRACT is specified. It cannot be specified when FORMAT=BACKUP.

Valid Entries:

DATACOM-NAME of a table in the database and area being loaded

Default Value:

(No default)

,TASKS=

Specifies the number of subtasks that can be used to update the index.

Valid Entries:

1 through 25

Default Value:

10