Submit - Build JCL Parameters Information

This section describes the options that are provided for building and submitting JCL.
cadadb219
This section describes the options that are provided for building and submitting JCL.
General Panel Information
The Submit-Build JCL Parameters panel contains the following sections:
  • Execution Options
    Controls how you submit the procedure.
  • Collection Options
    Controls how statistics are gathered for an extract procedure. Enter values for all these fields if you are executing an extract procedure.
  • Report Options
    Controls how the report procedure is submitted. Enter values for these fields if you are executing a report procedure.
  • Action Options
    Controls where the action procedure job stream is generated and specifies the model member names that are used for the generated job stream. Enter values for these fields if you are executing an extract procedure that is linked to an action procedure. The linked procedure causes
    Database Analyzer
    to generate a job stream. You are required to provide values for the Job and DSN fields. The pre- and post-action fields are optional.
For field descriptions, press PF1 to access online help.
Read-Ahead Buffer Recommendations
If read-ahead buffers are available, less I/O overhead is used.
Database Analyzer
also uses storage space above the line in XA systems for the read-ahead buffers. For full cylinder reads on tablespaces, the following values are recommended:
  • VSAM
    • Device Type -- 3390
    • Recommended Read-ahead Buffers -- 360
  • VSAM
    • Device Type -- 3380
    • Recommended Read-ahead Buffers -- 300
  • DIV
    • Device Type -- 3390
    • Recommended Read-ahead Buffers -- 1
  • DIV
    • Device Type -- 3380
    • Recommended Read-ahead Buffers -- 1
Pre-Action and Post-Action Models
You can append certain job steps before and after the job stream for an action, such as error condition traps. These job steps can be entered in any model members that you create. You can also append certain job steps before and after the job stream for a utility job.
To use the pre- and post-action job steps, specify the names of the model members containing the steps. Model member names can be entered from the Submit panels for procedures. Pre- and post-action job steps can be specified through the Utility Manager.
Pre-action job steps are included as the initial steps, and the post-action job steps are included after the generated action steps. If the number of steps that are generated in an action exceeds 200, the model member that is used to generate the first JOB statement is used to generate subsequent JOB statements. The following sample shows this processing order:
JOBCARD _ _ PRE-ACTION STEPS _ _ ACTIONS (for example, process objects #1-200) _ _ JOBCARD _ _ ACTIONS (for example, process objects #201-400) _ _ POST-ACTION STEPS
If an action procedure has conditions that are tied to it and all objects are rejected (for example, none of the objects meet the criteria), pre- or post-action JCL is not generated.
Example
You can create a model member that contains pre-action job steps. For example, you can define that the pre-action steps generate required accounting information.
For the post-action model member, you might define that the job steps include a COND parameter. For example, you would specify COND=ONLY to indicate that the specified step is executed only if a previous step abended.
Specify pre-action and post-action model members from the Submit-Build JCL Parameters panel. The pre-action job steps are added to the beginning of the job stream and the post-action job steps are added to the end of the job stream. Therefore, the pre-action model steps are generated first. The requested action or utilities are generated next. The post-action model steps are generated last.
When you specify the model members from the Submit-Build JCL Parameters panel, you can provide only one pre-action and one post-action model member. However, you can create as many pre- and post-action model members as you like. For example, you might create another post-action member whose job steps inform all users when the requested action or utilities are completed.
You are not prompted for user symbolics within the pre- and post-action model members.
Symbolic variables that are Db2 object-specific, such as %OBJECT, %CREATOR, %DBNAME, %STEPNAME (for post-action), %TSNAME, %PLANS, and %PKGS, are not supported in pre- and post-action model members, even if
Database Analyzer
can try to process them in a limited fashion. These symbolic variables are not supported because the pre- and post-actions are generated at the beginning and end of an action job stream. When the action is tied to many objects in an extract,
Database Analyzer
cannot determine which object to select to substitute for the object specific symbolic variables.
To use the %DBNAME symbolic variable in pre- and post-action model members, try using only one database in the extract. However, object-related symbolics must not be used in the pre- and post-action model members.
Scan Options
The Collection Type field in the Execution Options section of the Submit - Build JCL Parameters panel specifies how statistics are retrieved during statistics collection.
Regular Scan Type
  • Collects
    All statistics from all extract procedures that are contained within execution procedures.
  • Reported On
    Reports available from Query Facility or Report Procedures Facility.
Fast Scan Type
  • Collects
    All statistics from all extract procedures contained within execution procedures.
    CPU intensive audits are bypassed by the fast scan (but are not bypassed by the regular scan).
  • Reported On
    Reports available from Query Facility or Report Procedures Facility.
The fast scan option is specified in the Collection Type field. Fast scan collects all statistics from all extract procedures that are contained within execution procedures, but performs no time audits. Fast scan offers performance gains over a regular scan under the following conditions:
  • The number of pages within the tablespace is more than 10,000.
  • An edit procedure (EDITPROC) is involved.
  • Db2 compression is used.
  • There are a large number of holes.
  • There are a large number of free MAPIDs.
The fast scan processing considerations are shown in the following list:
  • The PAGESAVE statistic is estimated based on proprietary sampling techniques. This is in contrast to a regular scan, which is more CPU intensive but 100% accurate. During a regular scan, every row is un-compressed, passed through the EDITPROC, or both to determine the exact PAGESAVE value.
  • Data pages are traversed only by record size, instead of traversing in increments or by MAPIDs. However, the following audits for data pages are bypassed as follows:
    • The large hole chain is not traversed.
    • The number of MAPIDs in the datapage header is not verified.
    • The page number is not checked to be sequential to the previous page.
    • The MAPID free chain is not traversed.
    • The free space amount is not verified.
    • The displacement that is specified by each MAPID is not verified.
    • The record type pointed to by each MAPID is not verified.
    • The MAPID within a record is not validated.
    • The record length is not validated against the catalog.
    In most cases, you can run fast scans for daily processing and can run a regular scan every weekend to verify the previous audits.
Space Map Scan Type
  • Collects
    Statistics from space map pages.
  • Reported On
    Reports available from the following:
    • Allocation Analysis Report
    • Volume Analysis Report
    • Page Update Report
Access the space map scan option from the Collection Type field. The space map scan option requests a scan of only the space map pages to collect statistics.
Use this scan to examine the statistics that are typically not collected for space map pages. Because only the space map pages are scanned, a space map scan is faster than a fast scan.
A space map scan ignores any other pages for statistical collection. Therefore, statistics that cannot be calculated from only the space map page are not collected.
All product reports are affected by the space map scan, but not all statistics can be gathered from only the space map pages. Therefore, some reports can only display a partial collection of the statistics that are gathered during a space map scan.
For index space map scans, %RO symbolics (using RO utility code) are low because the number of keys in that index cannot be calculated. This does
not
include ROHIALLC and ROHIUSED. If an IBM Reorg Index is required, a fast scan is recommended.
All Statistics
The following tablespace/table reports can display information relating to the space map scans for all fields:
  • Allocation analysis
  • Volume analysis
  • Page update
  • Dropped tables (
    Database Analyzer
    versus Catalog)
  • New/dropped tables in Autobuild
  • Non-Db2 data sets on Db2 volumes
The following indexspace reports can display information relating to the space map scans for all fields:
  • Allocation analysis
  • Volume analysis
  • Dropped tables (
    Database Analyzer
    versus Catalog)
  • New/dropped indexes in autobuild
  • Non-Db2 data sets on Db2 volumes
Partial Statistics
Not all statistics can be gathered from only the space map pages. Some reports can only display a partial collection of the statistics that are gathered during a space map scan.
The following tablespace/table reports display some or partial space map scanned statistics:
  • Table Row Analysis
    Statistics are available only for Pages Found and Total Rows.
    For a non-segmented tablespace, Total Rows is set to 10,000. For a segmented tablespace, Total Rows is calculated as the number of used pages x 4074 divided by the average table row length.
  • Space Summary
    All statistics that are listed under the Space Allocation column are available; only the # of Tables statistics is available under the Space Analysis column. Most statistics are available under the Page Analysis column. However, estimates are returned for the Row Mean, Reorg Estimate (REORGP), and the %C values because row statistics are not available.
  • Freespace Volume Analysis
    Only statistics for Freespace are unavailable.
  • Page Freespace Analysis
    Only statistics for Used Pages are available.
  • Freespace Summary
    Only the statistics for Total Freespace statistic are available.
  • Rowlength Distribution
    Only the statistics for the Distribution of Row Lengths are unavailable.
The following indexspace reports display some or partial space map scanned statistics:
  • Freespace Volume Analysis
    Only statistics for Freespace are unavailable.
  • Page Freespace Analysis
    Only statistics for Active Pages are available.
  • Page Detail Analysis
    Only Total Pages, Active Pages, and Reorg Pages are available.
  • Space Summary
    All statistics that are listed under the Space Allocation column are available. Only the key length and attributes are available under the Space Analysis column. The leaf, non-leaf, and Average leaf distance statistics are not available. Because row statistics are not available, estimates are returned for Reorg Estimate (REORGP) and the %C values.
No Statistics Available
The following reports are not available because they display row, column, leaf, or non-leaf statistics that are not available during a space map scan:
  • Tablespace and table reports:
    • Hole Analysis
    • Relocated Row Distance
    • Relocated Row Distribution
    • Table Column Detail
    • Table Column Values
    • Table Row Analysis
    • Table Row Distribution
    • Table Row Freespace Distribution
    • Table Column Frequency Values
  • Indexspace reports:
    • Clustering Row Analysis
    • Column Detail Analysis
    • Column Value Analysis
    • Dropped Indexes -
      Database Analyzer
      vs. Catalog
    • Index Column Frequency Values
    • Leaf Distance Analysis
    • Ordering Key Analysis