Back To LeighWeb Mainframe Utilities Page

 ESFRECD is a utility design to streamline the process of reflecting changes in
 DB2 data structures in the CSP records which point to those data structures.

 It's initial premise or raison detre is that DBA's typically only deal with CSP
 and ENDEVOR in the area of CSP records.  The process of logging into CSP in
 CICS and ENDEVOR in TSO if one is not usually in these environments and will
 only be there briefly can be tedious.  Additionally, the issue of MSL
 maintenance is unnecessarily added to the equation.  Since the DB2 catalog
 contained everything needed to update the CSP record and DB2, CSP and ENDEVOR
 all can be updated in batch, this utility was created to automate those steps
 into a batch job.

 ESFRECD performs this CSP record update process in the following steps:

  1. Create a temporary MSL and EZEWORK dataset.  This insures that no conflict
     will arise with the user's own MSL and EZEWORK dataset.  It also insures
     that "stray" CSP ITEMs will not mess things up unexpectedly.

  2. Initialize the temporary MSL with batch CSP so that ENDEVOR can write to
     it.

  3. Retrieve the record element from ENDEVOR into the temporary MSL.

  4. Export the record from the temporary MSL into a sequential file in ESF
     format.

  5. Extract the column information from DB2 which is necessary to refresh
     (actually rebuild) the CSP record.

  6. Sort this information in column sequence order and remove some special HEX
     characters which will give subsequent batch TSO steps some problems.

  7. Invoke batch TSO/ISPF to create an ISPF table of the DB2 table information
     and an ISPF table of the prologue information for the record from either:

     A. The prologue "override" dataset created by using a "Y", "E", or "R" in
        the "CSP Prologue Override" field.
     B. The ESF CSP record prologue from step #4 above.
     C. If not present, from the DB2 REMARKS column for that table/view.

     Once these ISPF tables are built, they are used in ISPF file tailoring
     services to create the new CSP record in ESF format.

  8. Delete/Define the temporary MSL again to insure you are starting with a
     clean MSL.

  9. Import the newly created ESF CSP record into the temporary MSL.

 10. Invoke ENDEVOR to ADD/UPDATE the CSP record back into ENDEVOR.

 When you invoke ESFRECD, you are presented with a screen on which you can type
 in the necessary information.  The data needed is as follows:

  1. A DB2 table or view name.  This name (unqualified) will the CSP record will
     point to.

  2. A DB2 CREATOR.  Even though the record will be unqualified in CSP, the
     qualification is needed to access the appropriate information in the DB2
     catalog table SYSCOLUMNS.
1
  3. A CSP record name.

  4. An ENDEVOR CCID.  Required for the ENDEVOR steps.

  5. An ENDEVOR comment.  Required for the ENDEVOR steps.

  7. A value in the "CSP Prologue Override" field.  This value must be one of
     the following:

     "N" - This indicates that the prologue from ENDEVOR or from DB2 should not
           be overridden in favor of a user-developed prologue.  This is the
           default value.

     "Y" - This indicates that the prologue from ENDEVOR or from DB2 should be
           overridden in favor of a user-developed prologue.  When this option
           is selected, ESFRECD determines if a prologue override dataset with
           the name userid.TEMP.ESFRECD.PROVRDSN exists.  If it does exist,
           nothing is done and the flag is just set to "Y".  If it does not
           exist, it is created (always on work packs) , and the prologue for
           the record is extracted from the CSP to DB2 cross reference file and
           placed in the file just created.

     "R" - This value indicates that the prologue override dataset is to be
           deleted and recreated using the same processing as an initial "Y"
           value described above.  After the refresh has taken place, the value
           of the flag becomes "Y".

     "E" - This indicates that the user wants to edit the override dataset to
           either view its contents and/or to change the contents.  If the
           prologue override dataset does not exist, it will be created and
           populated like the Refresh or initial "Y" instance described above,
           and then edited.  If it does exist, it will just be edited
           immediately.  When the edit session ends, the flag value becomes "Y".

  8. A Y/N flag to trigger ESFRECD to generate the batch job.

  9. A Y/N flag to indicate whether the generated job should be edited, or not
     edited and submitted automatically.

 Two special features are supported from the panel to present a list of DB2
 tables/views to choose from and to access a CSP/DB2 cross reference file to
 populate the CSP record field if this is an existing record.

 If you can't remember how to spell a given table/view or just which one you
 want to choose, specify the table name (and/or the CREATOR) with a '%' at the
 end of the string you DO know, and ESFRECD will query the DB2 catalog using a
 "LIKE" WHERE predicate and bring up a panel with a list of tables/views to
 choose from.  On this panel, type an "S" next to the table/view name you wish
 to choose and press ENTER.  You are returned to the main ESFRECD panel and the
 TABLE and CREATOR fields are populated with your choice.
1
 If this is an existing CSP record, you now need to know which record name is
 associated with this DB2 object.  On UNISTAR, we have been maintaining a
 cross-reference file of CSP record names and DB2 table/view names.  It is in
 'DUSC1.STR.DOCLIB(CSPDB2XR)' and is recreated each time the CSPDB2XR utility is
 executed.  Since its format is consistant, ESFRECD can reference it and extract
 the information it needs.  To use the cross reference to determine the CSP
 record name, you must have the DB2 table/view name field populated with a full
 name and a "?" in the CSP record name.  When you press  with these
 fields set in this manner ESFRECD will access the cross-reference file and
 return the name of the CSP record into the CSP record field on the panel.

 Now you're ready to generate a job to actually do the work!

 To do this, just change the "N" to a "Y" in the "Generate JCL Now?" field and
 press .  If the "Review JCL Prior To Submit ?" field value is "N", the
 generated job will be submitted automatically.  If the value is "Y", however,
 you will be taken into an edit session on the JCL.  You must submit the JCL
 yourself if you choose this route.