Installation Procedure

This section describes the procedure for Adabas installation in z/VSE environments.

• Installation Checklist

• Contents of the Release Tape

• Preparing to Install Adabas

• Initializing the Adabas Communication Environment

• Installing the Adabas Database

• Migrating an Existing Database

• Logical Unit Requirements

• Job Exit Utility

• Acquiring Storage for the ID Table

• Acquiring Storage for the IIBS Table

• SVC Work Areas

• Displaying Storage Allocation Totals

• Calls from Other Partitions

• Dummy Sequential Files

• Backward Processing of Tapes and Cartridges

• Applying Zaps (Fixes)

• Adabas 8 Adalink Considerations

• Setting Defaults in ADARUN

Installation Checklist

The following is an overview of the steps for installing Adabas on a z/VSE system.

Contents of the Release Tape

The following table describes most of the libraries included on the release (installation) medium. Once you have unloaded the libraries from the medium, you can change these names as required by your site, but the following lists the names that are delivered when you purchase Adabas for z/VSE environments.

Preparing to Install Adabas

The major steps in preparing for Adabas installation are

• checking for the correct prerequisite system configuration; and

• allocating disk and storage space.

The following sections describe the nominal disk and storage space requirements, and how to allocate the space.

• Disk Space Requirements for Libraries

• Disk Space Requirements for the Database

• Data Sets Required for UES Support

• Disk Space Requirements for Internal Product Data Sets

• Adabas Nucleus Partition/Address Space Requirements

• Defining the Library

• Restoring the ADAvrs LIBR File

• Using the ADAvrs LIBR File

Disk Space Requirements for Libraries

The Adabas library requires a minimum of 3390 disk space as shown below. A certain amount of extra free space has been added to the requirements for library maintenance purposes.

This space is needed for Adabas objects and phases as well as source and JCS samples.

Disk Space Requirements for the Database

The Adabas database size is based on user requirements. For more information, refer to Adabas DBA Tasks. Suggested sizes for an initial Adabas database, allowing for limited loading of user files and the installation of Natural, are as follows.

The minimum 3390 disk space requirements are:

Data Sets Required for UES Support

The Software AG internal product libraries (APS - porting platform) are required if you intend to enable a database for universal encoding service (UES) support. These libraries are now delivered separately from the product libraries.

For UES support, the following libraries must be loaded and included in the LIBDEF concatenation:

APSvrs.LIBR
APSvrs.L0nn

where vrs is the version of the library provided on the most recent installation medium for these components and aa is LD, LC, or LS and nn is the load library level. If the library with a higher level number is not a full replacement for the lower level load library(s), the library with the higher level must precede those with lower numbers in the LIBDEF concatenation.

Also for UES support, the following library must be loaded and included in the session execution JCL:

ADAvrsCS.LIBR

For information about setting up connections to UES-enabled databases, see section Enabling Universal Encoding Support (UES) for Your Adabas Nucleus.

Disk Space Requirements for Internal Product Data Sets

The minimum disk space requirements on a 3390 disk for the internal product libraries delivered with Adabas Version 8 are as follows:

Adabas Nucleus Partition/Address Space Requirements

The Adabas nucleus requires at least 900-1024 KB to operate. The size of the nucleus partition may need to be larger, depending on the ADARUN parameter settings. Parameter settings are determined by the user.

Defining the Library

It is necessary to define the library before restoration. The following two examples show how VSAM and non-VSAM libraries are defined.

Defining a VSAM Library

The following is a job for defining a VSAM library:

// JOB DEFINE DEFINE VSAM V8 ADABAS LIBRARY
// OPTION LOG
// EXEC IDCAMS,SIZE=AUTO
DEFINE CLUSTER -
(NAME(ADABAS.ADAvrs.LIBRARY) -
VOLUME(vvvvvv vvvvvv) -
NONINDEXED -
RECORDFORMAT(NOCIFORMAT) -
SHR(2) -
TRK(nnnnnn)) -
DATA (NAME(ADABAS.ADAvrs.LIBRARY.DATA))
/*
// OPTION STDLABEL=ADD
// DLBL SAGLIB,'ADABAS.ADAvrs.LIBRARY',,VSAM
// EXEC IESVCLUP,SIZE=AUTO
ADABAS.ADAvrs.LIBRARY
/*
// EXEC LIBR
DEFINE L=SAGLIB R=Y
DEFINE S=SAGLIB.ADAvrs REUSE=AUTO R=Y
LD L=SAGLIB OUTPUT=STATUS
/*
/&

—where
vvvvvv vvvvvv are the volumes for primary and secondary space.
nnnnnn is the number of tracks for primary and secondary space.
vrs is the Adabas version.

Notes:

1. For FBA devices the tracks (TRK...) operand is replaced by the blocks (BLOCKS...) operand.
2. SAGLIB is the name of the Adabas library. The name SAGLIB can be changed to suit user requirements.

Defining a Non-VSAM Library

The following is a job for defining a non-VSAM library:

// JOB DEFINE DEFINE NON-VSAM V8 ADABAS LIBRARY
// OPTION LOG
// DLBL SAGLIB,'ADABAS.ADAvrs.LIBRARY',2099/365,SD
// EXTENT SYS010,vvvvvv,1,0,ssss,nnnn
// ASSGN SYS010,DISK,VOL=vvvvvv,SHR
// EXEC LIBR
DEFINE L=SAGLIB R=Y
DEFINE S=SAGLIB.ADAvrs REUSE=AUTO R=Y
LD L=SAGLIB OUTPUT=STATUS
/*
/&

where:
SYS010 is the logical unit for Adabas library.
vvvvvv is the volume for Adabas library.
ssss is the starting track or block for specified library.
nnnn is the number of tracks or blocks for specified library.
vrs is the Adabas version.

Restoring the ADAvrs LIBR File

Restore the ADAvrs LIBR file into sublibrary SAGLIB.ADAvrs. See the next section for information about preparing modules to run without the ESA option active.

If you have a license for one of the following Software AG products, restore the file into the appropriate sublibrary:

For information about installing these products, see the documentation for that product.

Using the ADAvrs LIBR File

Where applicable, modules for Adabas are shipped with AMODE=31 active.

Storage Above or Below the 16-MB Limit

Adabas can acquire storage above the 16-megabyte addressing limit. This capability allows Adabas to acquire the buffer pool (LBP), work pool (LWP), format pool (LFP), and attached buffers (NAB) above 16 MB.

Where applicable, modules for Adabas are shipped with AMODE=31 active. If you prefer to have buffers placed below the 16-megabyte limit, ADARUN must be relinked with AMODE=24.

User Program Execution in AMODE=31 and RMODE=ANY

Programs that will execute AMODE=31 or RMODE=ANY must be relinked with the new ADAUSER object module.

In addition, because the IBM VSE LOAD macro cannot be issued in RMODE=ANY, the IBM VSE CDLOAD macro must be used. Therefore, the zap to change the ADAUSER CDLOAD to the LOAD macro cannot be used.

Initializing the Adabas Communication Environment

Communication between the Adabas nucleus residing in a z/VSE partition and the user (either a batch job or TP monitor such as Com-plete or CICS) in another partition is handled with an Adabas SVC (supervisor call).

The program ADASIP is used to install the Adabas SVC. The system can run ADASIP to dynamically install the SVC without an IPL. Special instructions apply when using z/VSE with the Turbo Dispatcher as described in the next section below.

For information about messages or codes that occur during the installation, refer to the Adabas Messages and Codes documentation.

• Installing the Adabas SVC with Turbo Dispatcher Support

• ADASIP Processing

• Running ADASIP

• Finding an Unused SVC

• Loading a Secondary Adabas SVC

• ADASIP Execution Parameters

• ADASIP Runtime Display

Installing the Adabas SVC with Turbo Dispatcher Support

The Adabas SVC module supports the IBM z/VSE Turbo Dispatcher environment.

In a Turbo Dispatcher environment, the Adabas SVC runs in parallel mode when entered. Adabas processes multiple SVC calls made by users in parallel.

ADASIP Processing

To enable Turbo support, ADASIP installs a z/VSE first-level interrupt handler (ADASTUB) that screens all SVCs. When ADASTUB finds an Adabas SVC, it passes control directly to the Adabas SVC.

If your system is capable of running the Turbo Dispatcher and you do not want to run a particular SVC through the Turbo interface, you can set the UPSI flag V to 1 to exclude a particular SVC from use through the Turbo interface. See the ADASIP UPSI statement.

You can activate the ADABAS SVC with multiple CPUs active by specifying UPSI C. ADASIP will dynamically de-activate and re-activate the CPUs if required. If multiple CPUs are active and the UPSI C has not been specified, the following messages will be displayed:

ADASIP60 Only 1 CPU can be active during ADASIP
ADASIP79 Should we stop the CPUs? (yes/no)

Answering yes to this message will allow activation to occur; the CPUs will be dynamically de-activated and re-activated. Answering no will terminate ADASIP.

The ADASTUB module is installed only once per IPL process. On the first run of a successful ADASIP, the following set of messages are returned:

ADASIP63 ADASTUB Module Loaded at nnnnnnnn
ADASIP78 VSE Turbo Dispatcher Version nn
ADASIP69 Turbo Dispatcher Stub A C T I V E

When running ADASIP for subsequent Adabas SVC installations, the following message is displayed for information only:

ADASIP74 Info : Stub activated by previous ADASIP

When dynamically re-installing an Adabas SVC that was previously installed with Turbo Dispatcher support, execute a SET SDL for the Adabas SVC only. Do not execute the SET SDL for ADANCHOR a second time.

Running ADASIP

ADASIP requires a prior SET SDL for the SVC, and therefore must run in the BG partition. To install the Adabas SVC without an IPL, execute the following JCS in BG.

Notes:

1. When using the EPAT Tape Management System, EPAT must be initialized before running ADASIP.
2. At execution time, the ADASIP program determines if a printer is assigned to system logical unit SYSLST. If no printer is assigned, messages are written to SYSLOG instead of SYSLST.

For information about the ADASIP parameters, see the section ADASIP Execution.

To automatically install the Adabas SVC during each IPL, insert the following JCS (or its equivalent) into the ASI BG JCS procedure immediately before the START of the POWER partition where

Without Turbo Dispatcher Support

The following sample is available in member ADASIP.X:

// DLBL SAGLIB,'ADABAS.ADAvrs.LIBRARY'
// EXTENT SYS010,volume
// ASSGN SYS010,DISK,VOL=volume,SHR
// LIBDEF PHASE,SEARCH=SAGLIB.ADAvrs
SET SDL
ADASVCvr,SVA
/*
// OPTION SYSPARM='svc,suffix' SVC NUMBER
// UPSI 00000000 UPSI OPTIONS FOR ADASIP
// EXEC ADASIP,PARM='NRIDTES=nn'

With Turbo Dispatcher Support

The following sample is available in member ADASIPT.X:

// JOB ADASIPT INSTALL THE ADABAS SVC (TURBO)
// OPTION LOG,NOSYSDUMP
// DLBL SAGLIB,'ADABAS.ADAvrs.LIBRARY'
// EXTENT SYS010,volume
// ASSGN SYS010,DISK,VOL=volume,SHR
// LIBDEF PHASE,SEARCH=SAGLIB.ADAvrs
SET SDL
ADASVCvr,SVA
ADANCHOR,SVA
/*
// OPTION SYSPARM='svc,suffix' SVC NUMBER
// SETPFIX LIMIT=100K REQUIRED; SEE NOTE 2
// UPSI 00000000 UPSI OPTIONS FOR ADASIP
// EXEC ADASIP,PARM='NRIDTES=nn'

Notes:

1. A SETPFIX parameter is required with Turbo Dispatcher support to page fix ADASIP at certain points in its processing. A value of 100K should be adequate.
2. The SET SDL statement for ADANCHOR is required for Turbo Dispatcher support. This is in addition to the SET SDL statement for ADASVCvr.

Finding an Unused SVC

Adabas requires an entry in the z/VSE SVC table. To find an unused SVC, use one of the following methods:

Method 1

Set the S flag specified in the UPSI for ADASIP to create a list of used and unused SVCs in the z/VSE SVC table.

Method 2

Obtain a listing of the supervisor being used.

Using the assembler cross-reference, locate the label SVCTAB; this is the beginning of the z/VSE SVC table. The table contains a four-byte entry for each SVC between 0 and 150 (depending on the z/VSE version).

Locate an entry between 31 and 150 having a value of ERR21. This value indicates an unused SVC table entry. Use the entry number as input to ADASIP.

Loading a Secondary Adabas SVC

You can optionally specify a suffix to indicate the version of an SVC, as shown in the previous JCS examples. This allows you to run two different versions of the SVC. Before specifying a suffix, however, you must have previously linked the second version of the SVC. In addition, you must have performed a SET SDL operation on the new SVC's name (for example, ADASVCxx).

To optionally specify a different Adabas SVC using ADASIP, specify the SVC suffix (the last two bytes in the form, ADASVCxx), as follows, where xx is the two-byte suffix of the new SVC:

// OPTION SYSPARM='svc,xx'

ADASIP Execution Parameters

This section describes the ADASIP execution parameters.

• OPTION SYSPARM= Statement
• UPSI Statement
• NRIDTES PARM= Option
• REPLACE PARM= Option
• DMPDBID PARM= Option

Runtime Display

OPTION SYSPARM= Statement

An optional correction (zap) can be applied to the Adabas ADASIP program to insert the default SVC so that no SYSPARM need be specified. See the section Applying Zaps.

UPSI Statement

// UPSI DSxTVCGx

Setting the UPSI byte is the user's responsibility. If the UPSI byte is not set, the SVC installation executes normally.

The UPSI byte is used to select the following options:

NRIDTES PARM= Option

The size of the ID table default supports up to 10 Adabas targets. However, the ADASIP program will allow you to increase this number by using this new option of the PARM operand on the EXEC card. To increase the size of the ID table to nn entries, specify the following when executing ADASIP:

// EXEC ADASIP,PARM='NRIDTES=nn'

where nn is the number of databases to be supported. Refer to the section Acquiring Storage for the ID Table for information about calculating the correct value for nn.

REPLACE PARM= Option

Specifying REPLACE=N or NO will cause warning messages ADASIP80 and ADASIP81 to appear if the SVC has been previously installed. Specifying REPLACE=Y or YES replaces the current SVC regardless of any active targets. The default value is REPLACE=NO. No abbreviation of the REPLACE keyword is supported.

Warning: Setting the REPLACE parameter to YES should be done carefully. Replacing an SVC while your targets are running can produce unpredictable results.

If both the NRIDTES and REPLACE keywords are specified, they must be separated by a comma. For example:

//EXEC ADASIP,PARM='NRIDTES=10,REPLACE=YES'

DMPDBID PARM= Option

This ADASIP option allows snap dumps of the Adabas command queue for a specified database ID (DBID). The dump is written to SYSLST. The OPTION SYSPARM statement must specify the SVC number to perform the snap dump. For example, to perform a snap dump of the database 5 command queue, issue:

// OPTION SYSPARM='svc,suffix'
// EXEC ADASIP,PARM='DMPDBID=5'

ADASIP Runtime Display

When ADASIP is run, the ADASIP00 message displays the current system level.

ADASIP00 ...ADABAS V8 VSE SIP STARTED
SIP IS RUNNING UNDER VSE/systype-mode
ADASIP00 ... (yyyy-mm-dd, SM=sm-level, ZAP=zap-level)
ADASIP00 ... SIP IS RUNNING UNDER OSYS LEVEL Vnnn
ADASIP00 ... SIP IS LOADING ADABAS SVC LEVEL Vnnn
ADASIP00 ... ADASIP IS LOADING ADABAS SVC AMODE=amode

Installing the Adabas Database

This section describes installation of the Adabas database for z/VSE systems. Note that all applicable early warnings and other fixes must first be applied. For descriptions of any messages or codes that occur, refer to the Adabas Messages and Codes documentation.

• Installing the Release Tape

• Prepare the Installation Sample JCS for Editing

• Modify, Assemble, and Link the Adabas Options Table

• Catalog Procedures for Defining Libraries and the Database

• Database Installation Steps

Installing the Release Tape

Copy the data sets from the supplied installation medium to your disk before you perform the individual installation procedure for each component to be installed.

The way you copy the data sets depends on the installation method and the medium used:

• If you use System Maintenance Aid (SMA), refer to the copy job instructions provided in the System Maintenance Aid documentation.

• If you are not using SMA and want to copy the data sets from CD-ROM, refer to the README.TXT file on the CD-ROM.

• If you are not using SMA and want to copy the data sets from tape, follow the instructions in this section.

This section explains how to copy the data sets .LIBJ, .LIBR and .LICS from tape to disk. All other data sets can be installed directly from the tape.

• Step 1: Copy Data Set COPYTAPE.JOB to Disk
• Step 2: Modify COPYTAPE.JOB on Your Disk
• Step 3: Submit COPYTAPE.JOB

Step 1: Copy Data Set COPYTAPE.JOB to Disk

• Modify the following sample job according to your requirements:

  * $$ JOB JNM=LIBRCAT,CLASS=0,                                          +
  * $$ DISP=D,LDEST=(*,UID),SYSID=1                                      
  * $$ LST CLASS=A,DISP=D                                                
  // JOB LIBRCAT                                                         
  * *****************************************                            
  *     STORE COPYTAPE.JOB IN LIBRARY                                    
  * *****************************************
  // ASSGN SYS004,nnn 
  // MTC REW,SYS004                                                        
  // MTC FSF,SYS004,4                                                      
  ASSGN SYSIPT,SYS004                                                      
  // TLBL IJSYSIN,'COPYTAPE.JOB'                                           
  // EXEC LIBR,PARM='MSHP; ACC S=lib.sublib'
  /*
  // MTC REW,SYS004
  ASSGN SYSIPT,FEC                                                         
  /*                                                                       
  /&                                                                     
  * $$ EOJ

  where:

  nnn is the tape address, and
  lib.sublib is the library and sublibrary in which the data set COPYTAPE.JOB is to be stored.

• Execute the job to copy the data set COPYTAPE.JOB to disk.

  COPYTAPE.JOB contains the JCL required to copy the data sets .LIBJ, .LIBR and .LICS from tape to disk.

Step 2: Modify COPYTAPE.JOB on Your Disk

• Modify COPYTAPE.JOB according to your requirements and set the disk space parameters as appropriate.

Step 3: Submit COPYTAPE.JOB

• Execute COPYTAPE.JOB to copy the data sets .LIBJ, .LIBR and .LICS to your disk.

Prepare the Installation Sample JCS for Editing

The following sample installation job is available in member INSTALL.X.

Run the following job to load the installation samples:

* $$ JOB JNM=PUNINST,CLASS=A,DISP=D
* $$ LST CLASS=A,DISP=D
* $$ PUN CLASS=p,DISP=D
// JOB PUNINST INSTALL SAMPLES FOR ADABAS
// OPTION LOG
// DLBL SAGLIB,'ADABAS.ADAvrs.LIBRARY'
// EXTENT SYS010
// ASSGN SYS010,DISK,VOL=volume,SHR
// EXEC LIBR
ACCESS SUBLIB=SAGLIB.ADAvrs
PUNCH ADAPROC.X /* PROCS FOR FILE AND LIBRARY DEFINITIONS */
PUNCH ADAIOOAL.X /* ADABAS OPTIONS TABLE CUSTOMIZATION */
PUNCH ADASIP.X /* ADASIP JOB (NON-TURBO DISPATCHER) */
PUNCH ADASIPT.X /* ADASIP JOB (TURBO DISPATCHER) */
PUNCH ADAFRM.X /* SAMPLE ADAFRM JOB */
PUNCH ADADEF.X /* SAMPLE ADADEF JOB */
PUNCH ADALODE.X /* LOAD DEMO FILE EMPLOYEES */
PUNCH ADALODV.X /* LOAD DEMO FILE VEHICLES */
PUNCH ADALODM.X /* LOAD DEMO FILE MISC */
PUNCH ADALODP.X /* LOAD DEMO FILES PERSONNEL & LOB */
PUNCH ADANUC.X /* SAMPLE NUCLEUS STARTUP */
PUNCH ADAREP.X /* SAMPLE ADAREP JOB */
PUNCH ADAINPL.X /* SAMPLE NATINPL TO INSTALL AOS */
/*
/&
* $$ EOJ

where p is the output class for punch, volume is the specified volume for the Adabas library, and vrs is the Adabas version.

Once the selected members in the INSTALL job are within the local editor facility, the customization can begin.

Modify, Assemble, and Link the Adabas Options Table

Customize and run job ADAIOOAL to assemble and link the Adabas options table for installation customization.

The following describes the IORDOSO macro, which must be assembled and linked to the Adabas sublibrary as PHASE ADAOPD. The member X.ADAIOOAL shipped with Adabas can be used for this purpose.

• IORDOSO Macro Overview
• IORDOSO Macro Parameters

IORDOSO Macro Overview

The IORDOSO macro allows you to customize Adabas operation in the following areas:

• Loading phases;

• IDRC compaction support for 3480 and 3490 tape devices;

• Interfaces to z/VSE disk space managers such as DYNAM/D;

• Interfaces to z/VSE tape managers such as DYNAM/T

• An option controlling how the system writes to fixed block addressing (FBA) devices;

• An option to write printer (PRINT and DRUCK) files under either DTFPR or DTFDI control;

• GETVIS message printing;

• Optional job exit processing;

• Options for controlling the creation of z/VSE JCS with the Adabas Recovery Aid utility ADARAI;

• Sequential file processing under VSAM/SAM;

• Input device control with SYS000 assignment;

• Name of external sort program.

IORDOSO Macro Parameters

The following parameters can be set in using the IORDOSO macro.

CDLOAD

CDLOAD={ NO | YES } 
           

COMPACT

COMPACT={ NO | YES } 

DISKDEV

DISKDEV=devtype
           

Notes:

1. Adabas requires device type information when opening files. However, there may be situations where the device cannot be determined before the open without additional operations; for example, when a z/VSE Disk Space Manager or Tape Manager is active, or when using VSAM/SAM sequential files. Adabas also determines the block size to be used for sequential I/O areas by device type.
2. Valid disk device types are 3390, 9345 and FBA.

DISKMAN

DISKMAN={ NO | YES }

DISKSYS

DISKSYS=sysnum
           

DISKTYP

DISKTYP=text

DTFDI

DTFDI={ NO | YES }
           

FBAVRF

FBAVRF={ NO | YES }
           

GETMMSG

GETMMSG={ NO | YES } 
           

JBXEMSG

JBXEMSG={ NO| YES | PRT }

JBXIMSG

JBXIMSG={ NO | YES | PRT }

JOBEXIT

JOBEXIT={ NO | YES }

PFIXRIR

PFIXRIR={ NO | YES }

PRTDSYS

PRTDSYS={ sysnum | SYSLST } 

PRTDSYS=050
.
// ASSGN SYS050,PRINTER

PRTRSYS

PRTRSYS={ sysnum | SYS009 } 

RAIDASG

RAIDASG={ NO | YES } 

RAITASG

RAITASG={ NO | YES }

SORTPGM

SORTPGM={ sortpgm | SORT } 

SYS000O

SYS000O={ NO | YES }
           

Medium - SYS000  DTF Type

Card              DTFCD
Disk              DTFSD
Tape              DTFMT

TAPEDEV

TAPEDEV=devtype
           

Notes:

1. Adabas requires device type information when opening files. However, there may be situations where the device cannot be determined before the open without additional operations; for example, when a z/VSE Disk Space Manager or Tape Manager is active, or when using VSAM/SAM sequential files. Adabas also determines the block size to be used for sequential I/O areas by device type.
2. Valid tape device types are 2400, 3410, 3420, 3480 and 8809. For device types 3480, 3490, 3490E or 3590, specify TAPEDEV=3480.

TAPEMAN

TAPEMAN={ NO | YES }

TAPESYS

TAPESYS=sysnum

TAPETYP

TAPETYP=text

VSAMDEV

VSAMDEV=devtype

Notes:

1. Adabas requires device type information when opening files. However, there may be situations where the device cannot be determined before the open without additional operations; for example, when a z/VSE Disk Space Manager or Tape Manager is active, or when using VSAM/SAM sequential files. Adabas also determines the block size to be used for sequential I/O areas by device type.
2. Valid disk device types are 3390, 9345 and FBA.

VSAMSEQ

VSAMSEQ={ NO | YES }

VSAMSYS

VSAMSYS=sysnum

Additional Parameters Used for Internal Control Only

Three additional parameters are also available but are used only for internal control and should not be changed from their default settings unless otherwise specified by your Software AG technical support representative:

IORTRAC={NO | YES}
IORTSIZ={3000 | tablesize}
IORTTYP=(1... 14)(, opt1 ... opt14 ).

Catalog Procedures for Defining Libraries and the Database

The job ADAPROC is divided into two procedures:

• ADAVvLIB defining the library or libraries; and

• ADAVvFIL defining the database.

Customize and catalog the two procedures before placing them back in the procedure library. The following specific items must be customized:

• file IDs for the database and libraries;

• volumes for libraries and database files;

• space allocation for database files.

The Adabas DEMO database files include ASSO, DATA, WORK, TEMP, SORT, CLOG, and PLOG.

Database Installation Steps

Follow the steps outlined below to install a new Adabas database under z/VSE.

• Step 1. Allocate and format the DEMO database.
• Step 2. Define the global database characteristics.
• Step 3. Load the demonstration (demo) files.
• Step 4. Install the product license file.
• Step 5. Start the Adabas nucleus and test the Adabas communications.
• Step 6. Test Adabas partition communications.
• Step 7. Load the Adabas Online System, if used.
• Step 8. Terminate the Adabas nucleus.
• Step 9. Back up the database.
• Step 10. Insert the ADARUN defaults.
• Step 11. Install the required TP link routines for Adabas.

Notes:

1. For information about running ADADEF, ADAFRM ADALOD, ADAREP, and ADASAV in steps 1-3, 5, and 8 below, see the Adabas Utilities documentation.
2. For information about customizing the nucleus job and about starting, monitoring, controlling, and terminating the nucleus, see the Adabas Operations documentation.

Step 1. Allocate and format the DEMO database.

Customize and run the ADAFRM utility job to format the DEMO database areas. The following specific items must be customized:

• the Adabas SVC number, the database ID, and database device type(s);

• sizes of the data sets for each ADAFRM statement.

Step 2. Define the global database characteristics.

Customize and run the ADADEF utility job to define the global definition of the database. The following items must be customized:

• the Adabas SVC number, the database ID, and database device type(s);

• ADADEF parameters.

Step 3. Load the demonstration (demo) files.

Customize and run the job

• ADALODE to load the sample demo file EMPL;

• ADALODV to load the sample demo file VEHI;

• ADALODM to load the sample demo file MISC; and

• ADALODP to compress and load the sample Personnel (PERC) demo file and its associated LOB demo file.

  Note:
  The Personnel demo file must be installed on a UES-enabled database because it includes wide-character format (W) fields.

For each job, the following items must be customized:

• the Adabas SVC number, the database ID, and database device type(s);

• ADALOD parameters.

Step 4. Install the product license file.

The product license file is supplied on the individual customer installation media or separately via an e-mail attachment. If the license file is provided with the installation media, you can follow the instructions in this step to install the license file. If the license file is supplied via an e-mail attachment, you must first transfer the license to z/VSE, as described in Transferring a License File from PC to a z/VSE Host Using FTP and then you can install it, as described in this step.

Installing the license file.

In z/VSE environments, the product license file can be installed either as a load module or as a library member.

To install the product license file as a module, complete the following steps:

1. Verify that the license file is stored in an Adabas source library or sequential data set (with RECFM=F or FB and LRECL=80), taking care to preserve its format as ASCII.

2. If you loaded your Adabas license file to a library, review and modify sample JCS job ASMLICAL.X, adjusting the library and volume specifications as appropriate for your site. If you loaded your Adabas license file to a data set, use sample ASMLICAV.X instead.

   Note:
   In sample jobs ASMLICAL.X and ASMLICAV.X, the standard label area is assumed to contain label information for library USERLIB. You can change this as appropriate for your library.

3. Submit modified sample job ASMLICAL.X or ASMLICAV.X.

   These sample jobs generate your Adabas license in ADALIC.PHASE. They assume that ADALIC.PHASE will be in a user sublibrary. If a user sublibrary is chosen for ADALIC.PHASE, this sublibrary must be included in the LIBDEF search chain in your Adabas nucleus startup JCS. You may find it more convenient to place ADALIC.PHASE directly into the Adabas ADAvrs sublibrary, to avoid the need to define additional libraries. During initial testing, Software AG recommends using a user sublibrary.

To install the product license file as a library member, complete the following steps:

1. Verify that the license file is stored in an Adabas source library (with RECFM=F or FB and LRECL=80), taking care to preserve its format as ASCII.

2. Make sure any previously created ADALIC load module is inaccessible in the Adabas load library being used by the nucleus jobs. Adabas first tries to load ADALIC and if unsuccessful it reads from DDLIC.

3. Provide all Adabas nucleus startup jobs with a DLBL statement in the following format:

   // DLBL DDLIC,'/libname/sublb/memname.memtype'

   where libname is the Librarian name of the library, sublib is the name of the sublibrary, memname is the license member name, and memtype is the license member type.

To install the product license file as a sequential data set, complete the following steps:

1. Verify that the license file is stored in a sequential file (with RECFM=F or FB and LRECL=80), taking care to preserve its format as ASCII.

2. Make sure any previously created ADALIC load module is inaccessible in the Adabas load library being used by the nucleus jobs. Adabas first tries to load ADALIC and, if unsuccessful, it reads from DDLIC.

3. Provide all Adabas nucleus startup jobs with DLBL, EXTENT and ASSGN statements in the following format:

   // DLBL DDLIC,'adabas.license.file'
   // EXTENT SYSnnn
   // ASSGN SYSnnn,DISK,VOL=volser,SHR

   where adabas.license.file is the physical file name, nnn is an unused logical unit, and volser is the volume serial on which the license file resides.

Step 5. Start the Adabas nucleus and test the Adabas communications.

Customize and run the job ADANUC to start up the Adabas nucleus. The following items must be customized:

• The Adabas SVC number, the database ID, and device type(s);

  Note:
  Be sure to include appropriate LIBDEF references for user sublibraries, especially the library containing the ADALIC license file. The licensing component MLCvrs must also be added to the LIBDEF SEARCH chain for load modules. These additional sublibraries can be added via the ADAVvFIL procedure, as required.

• ADANUC parameters.

Step 6. Test Adabas partition communications.

Customize and run the job ADAREP in MULTI mode with the CPLIST parameter to test Adabas partition communications. The following items must be customized:

• the Adabas SVC number, the database ID, and device type(s);

• ADAREP parameters.

Step 7. Load the Adabas Online System, if used.

Customize and run the job ADAINPL to load the Adabas Online System into a Natural system file. A Natural file must first be created, requiring an INPL input file (see the Natural installation instructions). The following items must be customized:

• the Adabas SVC number, the database ID, and device type(s);

• the Natural INPL parameters and system file number.

Step 8. Terminate the Adabas nucleus.

Communicate with the Adabas nucleus (MSG Fn) to terminate the session by entering the Adabas operator command ADAEND into the Adabas nucleus partition.

Step 9. Back up the database.

Customize and run the ADASAV utility job to back up the Version sample database. The following items must be customized:

• the Adabas SVC number, the database ID, and device type(s);

• ADASAV parameters.

Step 10. Insert the ADARUN defaults.

Optionally customize and run the DEFAULTS job to set the ADARUN defaults using the MSHP utility and to relink ADARUN. The following items may be customized:

• SVC number;

• database ID;

• device type(s).

Step 11. Install the required TP link routines for Adabas.

Refer to the section Installing Adabas With TP Monitors for the TP link routine procedure.

Migrating an Existing Database

Use the ADACNV utility to migrate existing databases to new releases of Adabas. See the Adabas Utilities documentation for more information.

Logical Unit Requirements

This section describes the Adabas logical unit requirements.

ADARUN

Utility

Nucleus

The highest logical unit used is SYS038 for the ADASAV utility. The programmer logical units default is described in the section Device and File Considerations. The system programmer should review these requirements to ensure that there are enough programmer logical units to run the desired utilities in the desired partitions.

Job Exit Utility

Adabas provides a job exit to perform two different functions:

• Librarian input override processing

  The exit scans a job stream for Librarian input override statements. These statements indicate that card input (ADARUN CARD or utility KARTE statements) for a job step is to come from Librarian members rather than from SYSRDR or SYSIPT.

• ADARAI JCS capture processing

  The exit captures JCS before it is modified by tape or disk management systems for later use by ADARAI.

You can set the job exit to perform either function or both. By default, the job exit performs Librarian input override processing.

This section covers the following topics:

• Installation and Initialization

• Librarian Input Override Processing

• Activating Adabas Use of Job Exit Processing

• Using the Job Exit Utility for ADARAI JCS Capture

• Job Exit Storage Requirements

• Optional Console or Printer Messages

• Diagnostic Functions

Installation and Initialization

The job exit can be installed during ASI processing or at any time afterward. It is installed in two steps:

to install the job exit:

1. Install programs SAGJBXT and SAGIPT in the SVA.

2. Run program SAGINST to initiate job exit processing.

You can include SAGJBXT in the $JOBEXIT list of eligible exits, but you must still place SAGIPT in the SVA and run SAGINST to allocate the required table(s).

SAGIPT runs above the 16-megabyte line if an appropriate 31-bit PSIZE is available. In addition, the table that stores information from input-override statements and/or the table that stores JCS for ADARAI use is placed in 31-bit GETVIS, if available.

SAGINST reads an input parameter that tells it whether to install the Librarian input override processing, ADARAI JCS capture processing, or both. The following parameter values are valid:

PARM=ADALIB (the default) installs Librarian input override processing
PARM=ADARAI installs ADARAI JCS capture processing

The following sample job control initializes the job exit:

* $$ JOB JNM=SAGEXIT,CLASS=0
* $$ LST CLASS=A,DISP=D
// JOB SAGEXIT
// LIBDEF *,SEARCH=SAGLIB.ADAvrs
// EXEC PROC=ADAVvLIB
SET SDL
SAGJBXT,SVA
SAGIPT,SVA
/*
// EXEC SAGINST,PARM=ADARAI,ADALIB
/&
* $$ EOJ

where vrs is the Adabas version.

Librarian Input Override Processing

If Librarian input override processing is specified, the job exit scans a job stream for input override statements indicating that card input (ADARUN CARD or utility KARTE statements) for a job step is to come from Librarian members rather than from SYSRDR or SYSIPT. By default, the exit can store a maximum of 2000 input override cards simultaneously throughout the system. Adabas uses this facility when processing CARD and KARTE parameters.

Enable Librarian input override processing by adding * SAGUSER control statements to the job control stream between the // JOB and // EXEC statements.

A * SAGUSER statement can have three keyword parameters: FILE, LIBRARY, and MEMBER.

FILE={ CARD | KARTE }

LIBRARY={ library.sublibrary | libdef.source }

MEMBER=name [,{ type | A } ]

The following is an example of a * SAGUSER control statement that specifies an alternate job exit member:

* SAGUSER FILE=CARD,MEMBER=NUC151

In the example above, Adabas searches the current libdef.source chain for member NUC151 with type A. If NUC151 is found, Adabas uses its contents as the nucleus startup parameters instead of SYSIPT.

To permit flexible startup processing, multiple SAGUSER statements may be specified for each file. In the following example, Adabas reads the input parameters first in member NUC151, then in member IGNDIB:

* SAGUSER FILE=CARD,MEMBER=NUC151
* SAGUSER FILE=CARD,MEMBER=IGNDIB

The following examples show the use of the LIBRARY parameter, and apply to z/VSE systems only:

* SAGUSER FILE=CARD,MEMBER=NUC151,LIBRARY=SAGULIB.TESTSRC

In the example above, Adabas searches sublibrary TESTSRC in the SAGULIB library for member NUC151 with type A. If NUC151 is not found in sublibrary TESTSRC of library SAGULIB, no further search is made. The DLBL and EXTENT information for the SAGULIB library must be available.

* SAGUSER FILE=CARD,MEMBER=NUC151.ADARUN,LIBRARY=SAGULIB.TESTSRC

In the example above, Adabas searches sublibrary TESTSRC in the SAGULIB library at nucleus initialization for member NUC151 with type ADARUN. The library member types PROC, OBJ, PHASE, and DUMP are not permitted.

Activating Adabas Use of Job Exit Processing

Specify JOBEXIT=YES to allow Adabas to use SAGUSER statements in the job stream and recatalog the Adabas options table (ADAOPD).

Using the Job Exit Utility for ADARAI JCS Capture

Once the job exit utility has been installed for ADARAI, all utilities that write information to the RLOG automatically obtain file information from the ADARAI table that the job exit maintains. Manual intervention is not required.

Job Exit Storage Requirements

The job exit requires from 84 to 298 kilobytes (KB) of SVA storage, depending on whether the Librarian input override interface and/or the ADARAI JCS interface is installed. Of that total,

• 2 kilobytes are used for program storage (PSIZE);

• 82-kilobyte GETVIS for Librarian input override storage; and

• 214-kilobyte GETVIS for ADARAI JCS storage.

When running in z/VSE on z/VSE hardware, all of the GETVIS and 1 kilobyte of the PSIZE can be run above the 16-megabyte line.

Optional Console or Printer Messages

You have the option of displaying, printing, or preventing these messages by specifying the JBXEMSG and JBXIMSG parameters in the Adabas options table.

Diagnostic Functions

After the job exit is installed, you can produce dumps of the two tables for diagnostic purposes. Executing SAGINST with the ADASIP UPSI statement:

• UPSI 10000000 produces a dump of the Librarian input override table;

• UPSI 01000000 produces a dump of the ADARAI JCS table.

If the size of these two tables needs to be changed for any reason, SAGIPT may be zapped before being loaded into the SDL:

• The Librarian input override table size may be changed from the default of X‘00014874' (84,084 bytes) to an appropriate value by zapping location X‘18'. When altering the SAGIPT.OBJ module, ESDID=002 is required on the MSHP AFFECTS statement.

• The ADARAI JCS table size may be changed from the default of X‘000355D6' to an appropriate value by zapping location X‘0C'.

Each element in the Librarian input override table is 42 bytes in length. The default table size assumes 10 SAGUSER statements per file name, 10 file names, and 20 partitions, plus two extra unused entries. This number is an estimate of maximum concurrent residency; each statement is removed from the table after it is used.

Each element in the ADARAI JCS table is 91 bytes in length. The default table size accommodates 2400 entries with each DLBL, TLBL, or EXTENT statement requiring an entry in the table. Whenever a JOB statement is encountered, all entries for that partition (task ID) are cleared from the table.

Acquiring Storage for the ID Table

The SYSTEM GETVIS is used to acquire storage for the ID table (IDT). This storage is acquired using the ADASIP at SVC installation time. The size of storage in the SYSTEM GETVIS depends on the number of IDT entries specified using ADASIP. The default number of IDT entries (IDTEs) is 10. The size can be calculated as follows:

SIZE (in bytes) =1024 (IDT prefix) + 96 (IDT header) + (32 x number of IDTEs)
= 1024 + 96 + (32 x 10)
= 1024 + 96 + 320
= 1440 bytes

Also, additional SYSTEM GETVIS storage is acquired. This storage permits users to communicate from multiple address spaces when Adabas is not running in a shared partition. In this case, the following formula is used to calculate SYSTEM GETVIS:

SIZE (in bytes) = 192 (CQ header) + (192 x NC value) + (4352 x NAB value)

It may be necessary to increase the SVA size to meet these requirements. To do so, change the SVA operand in the appropriate $IPLxxx procedure, then re-IPL.

Acquiring Storage for the IIBS Table

The 31-bit SYSTEM GETVIS is used to acquire storage for the IIBS table (IIBS). This storage is acquired using the ADASIP at SVC installation time. The size of storage in the 31-bit SYSTEM GETVIS is 128K.

SVC Work Areas

For each Adabas SVC installed, a number of 384-byte work areas are reserved. The number of work areas reserved is calculated as four times the number of IDTEs (4 x IDTE-count). The maximum number of work areas allocated is 128; the minimum is 24. The SVC work areas therefore occupy between 9K and 48K of storage. The default value of 10 IDTEs results in 15K of SYSTEM GETVIS being allocated.

Displaying Storage Allocation Totals

Specifying // UPSI xxxxxxxGx during the ADASIP execution (see UPSI byte description in ADASIP Execution Parameters, earlier in this guide,) will generate allocation messages on the system console, showing the total 24-bit GETVIS and 31-bit GETVIS storage allocated by Adabas:

ADASIP85 GETVIS-24 storage allocated: nnnK
ADASIP85 GETVIS-31 storage allocated: nnnK

Calls from Other Partitions

In order for an Adabas nucleus to accept calls from other partitions, storage is acquired in the SVA GETVIS area for any required attached buffers. The buffers hold data moved between the nucleus and users in other partitions.

Dummy Sequential Files

If the file is not needed, it can be unassigned or assigned IGN such as the following:

// ASSGN SYS014,UA

or

// ASSGN SYS014,IGN

Backward Processing of Tapes and Cartridges

To perform backward processing of tapes or cartridges, file positioning must occur before the file is opened. This can only be done when an assignment is made for the file. When performing the ADARES BACKOUT utility function, the // ASSGN ... for file BACK must be done explicitly.

No tape management system can be used, because such systems perform the assign operation when the file is opened; the LUB and PUB remain unassigned until this occurs.

Applying Zaps (Fixes)

The jobs described in this section can be used to permanently change defaults and apply corrections (zaps) to the libraries in the supported z/VSE systems.

Two methods are used in z/VSE for applying corrective fixes to Adabas:

• the MSHP PATCH facility requires no definition of Adabas as a product/component on the MSHP history file. This method only alters phases. If the phase is relinked, the zap is lost.

• the MSHP CORRECT facility requires the definition of Adabas as a product/component using MSHP ARCHIVE.

Software AG distributes Adabas zaps to z/VSE users in MSHP CORRECT format and therefore recommends that you use MSHP CORRECT.

• Applying Fixes Using MSHP PATCH

• Applying Fixes Using MSHP CORRECT

• Link Book Update Requirements for Secondary SVC

• Link Book Update Requirements for Running AMODE 24

Applying Fixes Using MSHP PATCH

A sample job for applying a fix to Adabas using MSHP PATCH is as follows:

// JOB PATCH APPLY PATCH TO ADABAS
// OPTION LOG
// EXEC PROC=ADAVvLIB
// EXEC MSHP
PATCH SUBLIB=saglib.ADAvrs
AFFECTS PHASE=phasenam
ALTER offset vvvv : rrrr
/*
/&

where
vrs is the Adabas version
saglib is the Adabas library name in procedure ADAVvFIL
phasenam is the Adabas phase to be zapped
offset is the hexadecimal offset into the phase
vvvv is the verify data for the zap
rrrr is the replace data for the zap

Applying Fixes Using MSHP CORRECT

• MSHP ARCHIVE
• MSHP CORRECT

MSHP ARCHIVE

For new users or users with no requirement to maintain multiple versions of Adabas, the following sample job can be used to define Adabas to MSHP.

// JOB ARCHIVE ARCHIVE ADABAS
// OPTION LOG
// EXEC PROC=ADAVvLIB
// EXEC MSHP
ARCHIVE ADAvrs
COMPRISES 9001-ADA-00
RESOLVES 'SOFTWARE AG - ADABAS Vv.r'
ARCHIVE 9001-ADA-00-vrs
RESIDENCE PRODUCT=ADAvrs -
PRODUCTION=saglib.ADAvrs -
GENERATION=saglib.ADAvrs
/*
/&

—where
vrs is the Adabas version
saglib is the Adabas library name in procedure ADAVvFIL

A different MSHP history file must be used for each version and revision level of Adabas to which maintenance is applied.

To preserve the MSHP environment of an older version level of Adabas during an upgrade to a new version, it is necessary to create an additional MSHP history file for use by the new version.

The following sample MSHP job can be used to create an additional history file for a new version of Adabas and define Adabas to it.

// JOB ARCHIVE DEFINE HISTORY AND ARCHIVE ADABAS
// OPTION LOG
// EXEC PROC=ADAVvLIB
// ASSGN SYS020,DISK,VOL=volhis,SHR
// EXEC MSHP
CREATE HISTORY SYSTEM
DEFINE HISTORY SYSTEM EXTENT=start:numtrks -
UNIT=SYS020 -
ID='adabas.new.version.history.file'
ARCHIVE ADAvrs
COMPRISES 9001-ADA-00
RESOLVES 'SOFTWARE AG - ADABAS Vv.r'
ARCHIVE 9001-ADA-00-vrs
RESIDENCE PRODUCT=ADAvrs -
PRODUCTION=saglib.ADAvrs -
GENERATION=saglib.ADAvrs
/*
/&

—where
vrs is the Adabas version
volhis is the volume on which the Adabas Vvr history file resides
start is the start of the extent on which the Adabas Vvr history file resides
numtrks is the length of the extent on which the Adabas Vvr history file resides
adabas.new.version.history.file is the physical name of the Adabas Vvr history file
saglib is the Adabas library name in procedure ADAVvFIL

Once migration to the new version is complete, you can either

• continue to use the new history file to apply subsequent fixes; or

• delete the old version of Adabas from MSHP and merge the new version into the standard MSHP history file.

A sample MSHP job to remove an old version of Adabas is provided below.

// JOB REMOVE REMOVE OLD ADABAS
// OPTION LOG
// PAUSE ENSURE MSHP HISTORY FILE BACKUP HAS BEEN TAKEN
// EXEC MSHP
REMOVE ADAvrs
REMOVE 9001-ADA-00-vrs
/*
/&

—where vrs is the old Adabas version

A sample MSHP job to merge an additional history file for Adabas into the standard MSHP history file is provided below.

// JOB MERGE MERGE SEPARATE ADABAS INTO STANDARD HISTORY
// OPTION LOG
// PAUSE ENSURE MSHP HISTORY FILE BACKUPS HAVE BEEN TAKEN
// ASSGN SYS020,DISK,VOL=volhis,SHR
// EXEC MSHP
MERGE HISTORY AUX SYSTEM
DEFINE HISTORY AUX EXTENT=start:numtrks -
UNIT=SYS020 -
ID='adabas.new.version.history.file'
/*
/&

—where
volhis is the volume on which the Adabas Vvr history file resides
start is the start of the extent on which the Adabas Vvr history file resides
numtrks is the length of the extent on which the Adabas Vvr history file resides
adabas.new.version.history.file is the physical name of the Adabas Vvr history file

MSHP CORRECT

The MSHP CORRECT and UNDO jobs use the history file identified by label IJSYSHF in the z/VSE standard label area. If Adabas is maintained from a different MSHP history file, include the following label information in the CORRECT or UNDO job:

// DLBL IJSYSHF,'adabas.new.version.history.file'
// EXTENT SYSnnn
// ASSGN SYSnnn,DISK,VOL=volhis,SHR

—where
volhis is the volume on which the Adabas Vvr history file resides
nnn is the user-defined SYS number
adabas.new.version.history.file is the physical name of the Adabas Vvr history file

A sample of the use of MSHP CORRECT to install a fix to Adabas is provided below.

// JOB CORRECT APPLY ADABAS FIX
// OPTION LOG
// EXEC PROC=ADAVvLIB
// EXEC MSHP
CORRECT 9001-ADA-00-vrs : Axnnnnn
AFFECTS MODE=modname
ALTER offset vvvv : rrrr
INVOLVES LINK=lnkname
/*
/&

—where
vrs is the Adabas version
x is the Adabas component (for example, N for nucleus)
nnnnn is the Adabas fix number
modname is the Adabas object module to be zapped and then relinked
offset is the hexadecimal offset to the beginning of the zap
vvvv is the verify data for the zap
rrrr is the replace data for the zap
lnkname is the link book for the phase affected

The CORRECT job updates object and phase in a single job step using the link book feature of MSHP. The INVOLVES LINK= statement automatically invokes the linkage editor after the object module is updated.

For a zap applied with the INVOLVES LINK= statement, the following UNDO can be used to remove the fix from both object module and phase:

// EXEC MSHP
UNDO 9001-ADA-00-vrs : Axnnnnn
/*

where vrs is the Adabas version, x is the Adabas component (for example, N for nucleus), and nnnnn is the Adabas fix number.

Adabas provides a link book containing parameters for invoking the linkage editor for each Adabas phase. The name of each link book begins with "LNK" and the member type is "OBJ".

No link book is provided for module ADAOPD or for any other programs distributed in source form. Programs distributed in source form continue to be modified using assembly and link jobs.

If you choose not to take advantage of the link book facility, remove the INVOLVES LINK= statement from any zap before applying it. You can then run the linkage editor step to recreate the phase separately, as before.

This may be done to link a temporary version of a phase into a separate sublibrary for testing purposes. However, it is also possible to maintain a separate test version of Adabas modules by defining an additional z/VSE system history file. See Maintaining a Separate Test Environment in z/VSE.

Link Book Update Requirements for Secondary SVC

If you use the link book facility and require a non-standard SVC suffix (for example, if you relink the Adabas 8 SVC to phase ADASVC11), you must remember to update the link book for the SVC (LNKSVC.OBJ) to reflect the new phase name.

The link book provided for ADASVC81 is LNKSVC.OBJ. It contains the following:

PHASE ADASVC81,*,NOAUTO,SVA
MODE AMODE(31),RMODE(24)
INCLUDE SVCVSE
INCLUDE SVCCLU
ENTRY ADASVC

To set up an SVC with suffix -11, you would need to update the link book as follows:

// DLBL SAGLIB,'adabas.Vvrs.library'
// EXTENT SYS010
// ASSGN SYS010,DISK,VOL=volser,SHR
// EXEC LIBR
ACCESS SUBLIB=SAGLIB.ADAvrs
CATALOG LNKSVC.OBJ REPLACE=YES
PHASE ADASVC11,*,NOAUTO,SVA
MODE AMODE(31),RMODE(24)
INCLUDE SVCVSE
INCLUDE SVCCLU
ENTRY ADASVC
/+
/*

—where
vrs is the Adabas version
adabas.Vvrs.library is the physical name of the Adabas vrs library
volser is the volume on which the library resides

Link Book Update Requirements for Running AMODE 24

If you use the link book facility and require AMODE 24 versions of any modules linked by default as AMODE 31 (ADARUN, ADASVC74), you must update the corresponding link book (LNKRUN.OBJ, LNKSVC.OBJ) to remove the MODE statement.

This link book update can be made using a method similar to that described in the previous section for the SVC suffix update.

Adabas 8 Adalink Considerations

• Link Routine User Exit 1 (Pre-Command) and User Exit 2 (Post-Command)

• LNKUES for Data Conversion

• ADAUSER Considerations

Link Routine User Exit 1 (Pre-Command) and User Exit 2 (Post-Command)

A pre-command user exit and a post-command user exit may be linked with an Adalink routine:

• Link routine user exit 1, LUEXIT1, receives control before a command is passed to a target with the router 04 call.

  Note:
  Special commands emanating from utilities and from Adabas Online System are marked as physical calls. These calls must be bypassed in user exits. These calls have X‘04' in the first byte (TYPE field) of the command's Adabas control block (ACBX). LUEXIT1 must check this byte and return if it is set to X‘04'. Be sure to reset R15 to zero on return.

• Link routine user exit 2, LUEXIT2, receives control after a command has been completely processed by a target, the router, or by the Adalink itself.

At entry to the exit(s), the registers contain the following:

Any registers except register 15 that are modified by the user exits must be saved and restored; the address of a save area for this purpose is in register 13.

If at return from LUEXIT1, register 15 contains a value other than zero (0), the command is not sent to the target but is returned to the caller. The user exit should have set ACBXRSP to a non-zero value to indicate to the calling program that it has suppressed the command: response code 216 (ADARSP216) is reserved for this purpose.

User buffers can be used to pass data between the link routine user exits 1 and 2 and Adabas nucleus user exits 11 and 4. Refer to User Buffers for more information.

An Adalink routine can return the following non-zero response codes in ACBXRSP:

LNKUES for Data Conversion

The Adabas 8 standard batch ADALNK is delivered with UES (Universal Encoding Support). The LNKUES module, as well as the modules ASC2EBC and EBC2ASC, are linked into the standard batch ADALNK. LNKUES converts data in the Adabas buffers and byte-swaps, if necessary, depending on the data architecture of the caller.

LNKUES is called only on ADALNK request (X'1C') and reply (X'20') calls if the first byte of the communication ID contains X'01' and the second byte does not have the EBCDIC (X'04') bit set.

• For requests, LNKUES receives control before LUEXIT1.

• For replies, LNKUES receives control after LUEXIT2.

By default, two translation tables are linked into LNKUES/ADALNK:

• ASC2EBC: ASCII to EBCDIC translation; and

• EBC2ASC: EBCDIC to ASCII translation.

If you prefer to use the same translation tables that are used in Entire Net-Work:

• In ASC2EBC and EBC2ASC, change the COPY statements from UES2ASC and UES2EBC to NW2ASC and NW2EBC, respectively.

• Re-assemble the translation tables and re-link LNKUES/ADALNK.

Both the Adabas and Entire Net-Work translation table pairs are provided in the section Translation Tables. You may want to modify the translation tables or create your own translation table pair. Be sure to (re)assemble the translation tables and (re)link LNKUES/ADALNK.

Refer to the member LNKLNK.OBJ for the current link-edit control statements for linking the ADALNK.PHASE. The following is a sample job for (re)linking ADALNK with LNKUES and the translation tables:

*
// JOB ...
// EXEC PROC=
// LIBDEF *,SEARCH=(search-chain-library.sublib ...)
// LIBDEF PHASE,CATALOG=(lib.sublib)
 PHASE ADALNK,*
	MODE AMODE(31),RMODE(24)
	INCLUDE LNKVSE8
	INCLUDE LINKIND
	INCLUDE LNKGBLS
	INCLUDE LNKUES
	INCLUDE ASC2EBC
	INCLUDE EBC2ASC
	INCLUDE LNKDSL
	INCLUDE RTRVSE
	INCLUDE JNMVSE
	ENTRY ADABAS
// EXEC LNKEDT

The (re)linked ADALNK must be made available to Entire Net-Work. If you are calling Adabas 8 and you do not have the correct LNKUES/ADALNK module, Adabas produces unexpected results: response code 022 (ADARSP022), 253 (ADARSP253), etc.

ADAUSER Considerations

ADAUSER is a program that links the user to Adabas. It is specific to an operating system and is independent of release level and mode. It can be used in batch and in some TP environments.

ADAUSER contains the entry point ADABAS and should be linked with all user programs that call Adabas. No other programs containing the CSECT or entry point name ADABAS can be linked in these load phases.

On the first Adabas call, ADAUSER (CDLOAD) loads the latest version of ADARUN. This makes the calling process release-independent. Subsequent Adabas calls bypass ADARUN.

ADARUN processes its control statements. For the ADARUN setting PROGRAM=USER (the default), ADARUN loads the non-reentrant Adalink modules. To load a reentrant batch link routine, use the ADARUN parameter PROGRAM=RENTUSER. This makes the calling process mode-independent.

Setting Defaults in ADARUN

The member DEFAULTS.X is available for setting the ADARUN defaults.

DEFAULTS.X uses MSHP CORRECT to install the fix.

The distributed source library contains member ZAPOPT which also contains a zap in MSHP format to set the ADARUN defaults.