Usage

Utility Programs › IDMSDBAN › Usage

Usage

Input

Input to the IDMSDBAN utility consists of statements to control the utility processing.

Output

The IDMSDBAN utility generates message listings.

Execution mode

You can execute the IDMSDBAN utility in local mode only.

Usage considerations

Specifying dictionary name for SQL-defined databases

When processing an SQL-defined database, you must specify the dictionary name or catalog segment that IDMSDBAN connects to using the SYSIDMS DBNAME parameter (not the DICTNAME parameter).

Note: For more information about the SYSIDMS parameter file, see the CA IDMS Common Facilities Guide.

Subschema

The subschema that describes the database areas to be processed must include all member record types for all sets included in the subschema.

Processing DDLCAT, DDLCATX, and DDLCATLOD areas

When processing the DDLCAT and DDLCATX areas, you must use the PROCESS statement SUBSCHEMA parameter and the IDMSCATZ subschema name.

When processing the DDLCATLOD area you must use the PROCESS statement SUBSCHEMA parameter and the IDMSCATL subschema name.

Areas

By default, IDMSDBAN locks the areas to be processed. If IDMSDBAN attempts to lock an area that is already locked, the utility terminates. To prevent this, perform one of the following actions:

Either vary offline or vary to retrieval the areas to be processed before running IDMSDBAN.
Specify UNLOCKED in the PROCESS parameter statement.
Note: If you specify UNLOCKED and the database is updated by another application during IDMSDBAN processing, the statistics in reports 2, 3, 4, and 5 may be inaccurate, and the messages in reports 1 and 1A may be misleading.

Page ranges

The page ranges selected for processing must include all database pages that can contain owner or member records of any set(s), constraints, or indexes to be processed.

If you specify a set, constraint, or index with an owner or member record outside the specified pages, IDMSDBAN will terminate with an error.

Indexed sets

If an indexed set is specified for processing, processing must also be requested for all other indexed sets for which the SR8 records can occur on the same pages as the SR8 records for the specified set.

IDMSDBAN has two parts

IDMSDBN1 and IDMSDBN2.

Intermediate work file size

The size of the intermediate work files will vary depending on the size and complexity of the database being analyzed.

IDMSDBN1 will generate one output record for each pointer position in each record for each set being processed. The file can be written to tape or disk; however, to ensure adequate space, tape is recommended.
IDMSDBN2 uses an intermediate work file to hold records it is working on. For every record it receives from IDMSDBN1, it will produce up to two intermediate work records. If the OWNCHECK option has been specified, it is necessary for the utility to produce these records for each occurrence of a record participating in a chained set.

Tape management systems and IDMSDBAN work files

The SYS002 work file used by IDMSDBN2 is written to and read from repeatedly. If this file is placed on a tape, then your tape management system may prevent IDMSDBN2 from overwriting the file after it has written to the file the first time. If the tape management system installation defaults allow, specify a zero retention period for the work file and/or specify DISP=(NEW,DELETE) for the work file.

If problems are detected

If the messages issued by IDMSDBAN indicate that problems exist in the database, the database should not be updated in between the time of the IDMSDBAN run and the time that the problems are corrected.

IDMSDBN1

IDMSDBN1 sweeps each specified area. For each area, IDMSDBN1:

Collects statistics
- Detailed statistics for each area and record type.
- Summary statistics for user-defined chained and indexed sets.
Generates an intermediate work file
IDMSDBN1 generates input to IDMSDBN2.

Each record in the file represents a pointer in a set connection between two records. For example, one record might represent the next pointer between two member records in a user-defined set; another might represent the connection between the root of a variable-length record and the first fragment of the record.

Each record includes the database keys of the two connected records. The database key of the record that contains the pointer to the other record is the FROM database key. The database key of the record to which the first record points is the TO database key.

Output

Report	Required/ Optional	Description
#1: Messages	Required	Lists input parameters used in the run. Lists all messages issued by IDMSDBN1. The messages: Report errors in the parameter input Trace the processing of areas Define any unexpected conditions detected by the program
#2: Area Information	Optional	Provides detailed area statistics. The report includes histograms of space availability and of data records per page.
#3: Set Statistics	Optional	Presents summary set statistics.
#4: Record Information	Optional	Provides detailed statistics for each record type in each area being analyzed. The report includes a histogram of data records per page.

IDMSDBN2

As a performance enhancement for the IDMSDBN2 step, the SIZE=E99999999 parameter has been added to the various sorts that are executed. However, some sort packages do not allow the usage of the SIZE= parameter on their SORT statements. If your sort package does allow the specification of the SIZE= parameter, you can reduce the run time of IDMSDBN2 by coding SORTSIZE=ON in the jobstep's SYSIDMS input file.

IDMSDBN2 does the following:

Verifies set integrity. This is done in three steps.
1. Reading the intermediate work file from IDMSDBN1
2. Iteratively massaging and sorting the records
3. Creating chains
  A chain is a path that originates at a record located at a FROM database key and terminates at a record located at a TO database key.
  
  IDMSDBN2 concatenates chains until a closed loop is created. A loop is created when the record at the last TO database key matches the record at the first FROM database key.
  
  Each chain created by IDMSDBN2 is associated with a set. IDMSDBN2 verifies the integrity of sets by ensuring that each chain is complete and contains one and only one owner.
Collects detailed set statistics.

Output

Report	Required/ Optional	Description
#1A: Messages	Required	Lists all messages issued by IDMSDBN2. The messages: Trace the processing of the chain file. Report inconsistencies detected during chain processing.
#5: Set Analysis Information	Optional	Provides detailed statistics for each set type processed. Chained sets—The report includes histograms of chain length, of page changes, and of pages used to store the set. Sets of variable-length-record fragments—The report includes a histogram of pages used to store the set. Indexed sets owned by a user record—The report includes histograms of SR8 usage in set occurrences, of members in set occurrences, and of SR8 levels in set occurrences. (SR8 records are internal index records.)

Report

Required/ Optional

Description

#1A: Messages

Required

Lists all messages issued by IDMSDBN2. The messages:

Trace the processing of the chain file.
Report inconsistencies detected during chain processing.

#5: Set Analysis Information

Optional

Provides detailed statistics for each set type processed.

Chained sets—The report includes histograms of chain length, of page changes, and of pages used to store the set.
Sets of variable-length-record fragments—The report includes a histogram of pages used to store the set.
Indexed sets owned by a user record—The report includes histograms of SR8 usage in set occurrences, of members in set occurrences, and of SR8 levels in set occurrences. (SR8 records are internal index records.)