This section describes how to install the Natural interface to DB2 (in the remainder of this section also referred to as NDB) in the various environments supported.
The installation procedures contain a number of options that depend on the TP monitor being used as well as on other site requirements.
This section covers the following topics:
Notation vrs or vr: If used in the following document, the notation vrs or vr stands for the relevant version, release, system maintenance level numbers. For further information on product versions, see Version in the Glossary.
The installation of Software AG products is performed by installation jobs. These jobs are either created manually or generated by System Maintenance Aid (SMA).
For each step of the installation procedure described later in the
section Installing Natural for DB2, the job number of a job performing the
respective task is indicated. This job number refers to an installation job
generated by SMA. If you are not using SMA, an example job of the same number
is provided in the job library on the NDB installation tape; you must adapt
this example job to your requirements. Note that the job numbers on the tape
are preceded by a product code (for example, NDBI070).
For information on the use of Software AG's System Maintenance Aid for the installation process, refer to the System Maintenance Aid documentation.
Base Natural must be installed first; you cannot install Natural and Natural for DB2 at the same time.
The Software AG Editor must be installed (see Installing the Software AG Editor in the Natural Installation documentation).
If you want to use the DB2 DL/I batch support (DSNMTV01),
the Natural interface to DL/I is required.
Further product/version dependencies are specified under Natural and Other Software AG Products and Operating/Teleprocessing Systems Required in the current Natural Release Notes.
The installation tape contains the datasets listed in the table below. The sequence of the datasets is shown in the Report of Tape Creation which accompanies the installation tape.
| Dataset Name | Contents |
|---|---|
NDBvrs.SRCE |
NDB source modules |
NDBvrs.LOAD |
NDB load modules |
NDBvrs.INPL |
NDB utility programs in INPL format
|
NDBvrs.ERRN |
NDB error messages |
NDBvrs.JOBS |
NDB installation jobs |
The notation vrs in dataset names represents the version number of the product.
SMA を使用する場合は、『System Maintenance Aid』ドキュメントを参照してください（Natural ドキュメント CD の最新版に含まれています）。
SMA を使用しない場合は、次の指示に従ってください。
このセクションでは、次の方法について説明します。
データセット COPY.JOB をテープからディスクにコピーします。
ローカルの命名規則に一致するように、このデータセットを変更します。
このデータセットの JCL を使用して、すべてのデータセットをテープからディスクにコピーします。
複数の製品のデータセットがテープで配布されている場合は、配布されたすべての製品のデータセットをテープからディスクにアンロードするための JCL がデータセット COPY.JOB に含まれています。
その後で、各コンポーネントに対して個別のインストール手順を実行する必要があります。
データセット COPY.JOB（ラベル 2）には、他のすべての既存のデータセットをテープからディスクにアンロードするための JCL が含まれています。 COPY.JOB をアンロードするには、次のサンプル JCL を使用します。
//SAGTAPE JOB SAG,CLASS=1,MSGCLASS=X //* --------------------------------- //COPY EXEC PGM=IEBGENER //SYSUT1 DD DSN=COPY.JOB, // DISP=(OLD,PASS), // UNIT=(CASS,,DEFER), // VOL=(,RETAIN,SER=tape-volume), // LABEL=(2,SL) //SYSUT2 DD DSN=hilev.COPY.JOB, // DISP=(NEW,CATLG,DELETE), // UNIT=3390,VOL=SER=volume, // SPACE=(TRK,(1,1),RLSE), // DCB=*.SYSUT1 //SYSPRINT DD SYSOUT=* //SYSIN DD DUMMY //
プレースホルダの意味は次のとおりです。
hilev は有効なハイレベル修飾子です。
tape-volume はテープボリューム名です（T12345 など）。
volume はディスクボリューム名です。
ローカルの命名規則に一致するようにディスクの COPY.JOB を変更し、ディスクのスペースパラメータを次のように設定してからこのジョブを送信します。
HILEV に有効なハイレベル修飾子を設定します。
LOCATION にストレージの場所を設定します。
EXPDT に有効な失効日を設定します。
COPY.JOB を送信して、その他すべてのデータセットをテープからディスクにアンロードします。
This section describes how to install Natural for DB2 in various environments:
This section covers the following topics:
The following steps describe the procedure for installing the components of NDB that are common to all environments.
Allocate a PDS as DBRM (database request module) library. The size of this dataset and the number of directory entries depend on the particular site (5 tracks and 20 directory blocks must be adequate for most environments). The PDS must have a fixed-block record format and a record length of 80.
Any standard dataset name can be used for this DBRM library; however,
this installation procedure assumes that the name SAGLIB.DB2DBRM
is used.
Job I055, Step 1600
By executing a standard Natural batch job, this step generates the
assembly source for NDBIOMO from the member
NDBIOTM.
This batch job invokes the Natural program NDBGENI, which
is loaded with INPL during the base Natural installation.
NDBGENI contains the following two parameters, which you can
modify to meet your specific requirements:
the DB-environment parameter, which must be set to:
DB2V7 if you are running DB2 Version 7
DB2V8 if you are running DB2 Version 8 or higher
the number of parallel dynamic prepared DB2 statements.
NDBIOMO (see also
the relevant section in Internal
Handling of Dynamic Statements) performs the dynamic
access to DB2 and contains all necessary EXEC SQL statements. In
addition, it contains some special SQL statements which cannot be executed in
dynamic mode.
An output report is created by this job. Check the report for successful job completion. In addition, a condition code of 0 indicates normal completion.
Job I055, Step 1610
Precompile, assemble and link NDBIOMO.
Note:
The link-edit step receives a condition code of 4 because of
unresolved references for DSNHLI. This is normal and can be
ignored.
Job I055, Step 1630
If desired, change library names and plan name to meet site requirements.
Job I055, Steps 1640/1650 or 1660/1670 or 1675/1676
The NDB parameter module NDBPARM contains the macro
NDBPRM with parameters
specific to the Natural interface to DB2.
You can generally use the default values for all parameters. Modify only the values of the parameters whose default values do not suit your requirements.
The individual parameters are described in the section Parameter Module NDBPARM.
When the file server is not to be used:
Execute the Steps 1640 and 1650; the resulting parameter module is
again called NDBPARM.
When the file server is to be used:
Execute the Steps 1660 and 1670; the resulting additional parameter
module is called NDBPARMF.
When the file server uses the Software AG Editor buffer pool as
storage medium:
Execute the Steps 1675 and 1676; the resulting additional parameter
module is called NDBPARME.
Job I055, Step 1680
Link-edit the environment-independent NDB nucleus
NATGWDB2.
Verify that the INCLUDE cards refer to the corresponding
DD names for the load libraries.
Adapt your Natural parameter module NATPARM by adding
parameters specific to Natural for DB2 (see Natural Parameter Modification for
DB2) and reassemble NATPARM.
Natural for DB2 basically consists of:
An environment-independent nucleus, which can be shared by multiple environments.
Environment-dependent components, which must be linked to the appropriate Natural environment-dependent interface.
Modify the JCL used to link your Natural shared nucleus by adding the
following INCLUDE card:
INCLUDE SMALIB(NATGWDB2) |
Environment-independent NDB nucleus from Step 6 |
Modify the JCL used to link your Natural environment-dependent nucleus
by adding the following INCLUDE cards and the corresponding DD
statements:
INCLUDE SMALIB(NDBPARM) |
NDB parameter module created in Step 5 |
INCLUDE SMALIB(NDBIOMO) |
NDB I/O module created in Step 3 |
INCLUDE DSNLIB(DSNTIAR) |
SQL Error Message Module |
INCLUDE
xxxxxxxx(yyyyyyyy)
|
Environment-dependent DB2 Interface (see below) |
If you want to use the Natural File Server, include
SMALIB(NDBPARMF) or SMALIB(NDBPARME) instead of
SMALIB(NDBPARM); see also Step 5 above.
Depending on your environment(s), INCLUDE the appropriate
environment-specific language interface
yyyyyyyy in the library
xxxxxxxx as shown in the following
table:
| Interface | Library | Environment |
|---|---|---|
DSNALI |
DSNLIB |
Under TSO and in batch mode without running under the control of the DSN command processor (that is, with CAF). |
DSNRLI |
DSNLIB |
WLM (Workload Manager) stored procedure address space and Natural Development Server (NDV) (recommended). Also usable in TSO and batch environments. |
DSNELI |
DSNLIB |
Under TSO and in batch mode when running under the control of the DSN command processor. |
DSNCLI |
DFHLIB |
Under CICS |
| DFSLI000 | IMSLIB |
Under IMS TM (MPP and BMP) and in batch mode by using the DB2
DL/I batch support (DSNMTV01).
|
NDBCOM |
NDBLIB |
Under Com-plete. |
Note:
If you want to use NDB in various environments (that is, with
different TP monitors), you must repeat this step for each of these
environments.
Instead of link-editing your Natural nucleus in the way described above, you have the following alternatives:
If you do not use a Natural shared nucleus, all modules must be included in the link-edit of the Natural nucleus.
Remove NATGWDB2 from the link-edit of the Natural
shared nucleus and run it as a separate module with the mandatory entry name
NATGWDB2. You can modify the name of the module created in Step 6.
However, if you use a name different from NATGWDB2, this name must
be specified as an alias name in an
NTALIAS
macro entry of the Natural parameter module. This way of link-editing only
applies if the Natural Resolve CSTATIC Addresses feature (RCA) is used.
Include all modules in the link-edit job of a separate Natural
parameter module with the mandatory entry name CMPRMTB. The name
of the resulting module is arbitrary. This way of link-editing only applies if
an alternative parameter module (PARM profile
parameter) is used. If link-editing is done in this way, you can install NDB
without having to modify your Natural nucleus or driver.
If link-editing is done according to number [2] or [3], the following applies:
| TP Monitor | Requirement |
|---|---|
| CICS | The resulting module must be defined via a PPT entry or RDO.
PPT entry: DFHPPT TYPE=ENTRY,PROGRAM=module-name,PGMLANG=ASSEMBLER |
| Com-plete | The resulting module must be defined as
RESIDENTPAGE or reside in the LPA/(E)LPA.
|
Job I061, Step 1610
Before executing this step, change the CMWKF01 DD
statement to point to the NDBnnn.INPL
dataset.
In this step, the NDB system programs, maps and DDMs are loaded into
the Natural system files. The INPL job loads objects into the
Natural system libraries SYSDDM, SYSTEM and
SYSDB2.
The NDB system programs must be loaded into the
FNAT system file.
 |
Warning: Ensure that your newly created SYSDB2 library
contains all necessary Predict interface programs, which are loaded into
SYSDB2 when installing Predict (see the relevant Predict
documentation). |
Job I061, Step 1620
Before executing this step, change the CMWKF02 DD
statement to point to the NDBnnn.ERRN
dataset.
This step executes a batch Natural job that runs an error load program
by using the NDBnnn.ERRN dataset as
input. The ERRLODUS job loads error messages into the library
SYSERR in the FNAT system file.
The NDB error messages must be loaded into the
FNAT system file.
Job I070, Steps 1604,1606,1608,1610
Create server stubs to execute Natural stored procedures and Natural user-defined functions. Natural for DB2 (NDB) server stubs are interface modules between the DB2 database system and the Natural server. In order to execute Natural stored procedures and Natural user-defined functions, the server stub needs to be installed.
There are two types of server stub (vr in the stub name stands for the current product version and release numbers):
The NDB server stub (module NDBvrSRV,
Steps 1604 and 1606)
The server stub is used to execute Natural stored procedures and
Natural user-defined functions.
The IBM LE (Language Environment) runtime modules required must be
linked to the NDB server stub module. Use the CALL option of the linkage editor
and assign the LE runtime library as SYSLIB.
The NDB start server stub (module
NDBvrSTR, Steps 1608 and 1610)
The start server stub is used to start the Natural server
environment(s) explicitly.
The IBM LE (Language Environment) runtime modules required must be
linked to the NDB start server stub module. Use the CALL option of the linkage
editor and assign the LE runtime library as SYSLIB. Additionally,
include the modules NDBSTRP (delivered with NDB) and
NATCONFG (delivered with Natural) from
NDBnnn.LOAD and
NATnnn.LOAD.
The NDB server stubs are generated from the
NDBSTUB macro. You can
generally use the default values for all parameters. Modify only the values of
the parameters whose default values do not suit your requirements. The
individual parameters are described in the section
NDB Server
Stub.
The resulting load modules have to be placed into a steplib library of the JCL used to execute the DB2 stored procedure address space.
For DB2 UDB, each Natural stored procedure or Natural user-defined
function must be defined by a DB2 CREATE PROCEDURE or DB2
CREATE FUNCTION statement, where the name of the NDB server stub
module NDBvrSRV generated is specified
as EXTERNAL NAME.
Job I070, Step 1615
Bind DBRM ROUTINEN into a package. DBRM
ROUTINEN is contained in the collection
SAGNDBROUTINENPACK and delivered with NDB. NDB needs this
collection for accessing the DB2 catalog and retrieving the parameter
descriptions of Natural stored procedures and Natural user-defined
functions.
This section describes how to install Natural for DB2 in a CICS environment.
Ensure that your Natural/CICS thread size is large enough to contain
the DB2SIZE; if you use the Natural Tools for DB2, an additional
storage of 8 KB is required.
This section covers the following topics:
If you want fixed assignment of your transaction code to the DB2 plan,
add an additional entry to your RCT, or create a DB2ENTRY with RDO
if you have CICS TS installed.
Perform one the following alternative steps:
Modify your RCT as follows (for any other parameters, refer to the relevant DB2 literature by IBM):
DSNRCT TYPE=ENTRY,PLAN=plan-name,TXID=(transaction-ID)
For parameters, refer to the relevant CICS literature by IBM.
DEFINE DB2ENTRY
OVERTYPE TO MODIFY CICS RELEASE = 0530
CEDA DEFine DB2Entry( )
DB2Entry : DB2ENTR
Group : NCI
DEscription :
THREAD SELECTION ATTRIBUTES
TRansid : transaction-id
THREAD OPERATION ATTRIBUTES
ACcountrec : None None ! TXid ! TAsk ! Uow
AUTHId :
AUTHType : Userid Userid ! Opid ! Group ! Sign ! TErm
! TX
DRollback : Yes Yes ! No
PLAN : plan-name
PLANExitname :
PRIority : High High ! Equal ! Low
PROtectnum : 0005 0-2000
THREADLimit : 0005 0-2000
THREADWait : Pool Pool ! Yes ! No
The plan-name must be the same as the name used to create the NDB plan, see NDB Common Installation Steps.
If you want to perform plan selection by using the dynamic plan exit, perform the following steps:
Job I070, Step 1630
The sample exit NDBUEXT can be modified to use a default
plan name if none has been specified prior to the first SQL call. Review the
source code in the module NDBUEXT for details about specifying a
default plan name.
Optionally modify the source module NDBUEXT.
Precompile, assemble and link NDBUEXT for CICS.
Note:
This step receives a condition code of 4 because of an unresolved
external reference for DFHEAI0 and DFHEI1. This is
normal and can be ignored.
Job I075, Step 1640
The resulting module NDBUEXT must be linked to the CICS
load library and defined via a corresponding PPT entry or RDO.
PPT entry:
DFHPPT TYPE=ENTRY,PROGRAM=NDBUEXT,PGMLANG=ASSEMBLER
Perform one the following alternative steps:
Modify your RCT as follows (for any other parameters, refer to the relevant DB2 literature by IBM):
DSNRCT TYPE=POOL,PLNPGME=NDBUEXT,PLNEXIT=YES
The parameter PLNPGME must specify the same program as
the NAME statement of the link step above.
For parameters, refer to the relevant CICS literature by IBM.
DEFINE DB2ENTRY
OVERTYPE TO MODIFY CICS RELEASE = 0530
CEDA DEFine DB2Entry( )
DB2Entry : DB2ENTR
Group : NCI
DEscription :
THREAD SELECTION ATTRIBUTES
TRansid : transaction-id
THREAD OPERATION ATTRIBUTES
ACcountrec : None None ! TXid ! TAsk ! Uow
AUTHId :
AUTHType : Userid Userid ! Opid ! Group ! Sign ! TErm
! TX
DRollback : Yes Yes ! No
PLAN :
PLANExitname : NDBUEXT
PRIority : High High ! Equal ! Low
PROtectnum : 0005 0-2000
THREADLimit : 0005 0-2000
THREADWait : Pool Pool ! Yes ! No
The parameter PLANExitname must specify the
same program as the NAME statement of the link step above.
Alternatively or additionally, you can specify the plan exit program
NDBUEXT with the PLANExitname parameter of
POOL THREAD ATTRIBUTES of the DB2Conn resource definition of CICS
TS.
If you want to use the Natural File Server (VSAM), perform the following additional steps:
Job I008, Step 1610
Specify the size and the name of the VSAM RRDS that is to be used as the file server (see also Installing the File Server in Natural File Server for DB2).
Job I075, Step 1610
Specify the five input parameters required to format the file server dataset (see also Installing the File Server in Natural File Server for DB2).
Shown below are sample additional CICS table entries needed for the file server and for the DB2 components of Natural:
FCT entry:
CMFSERV DFHFCT TYPE=DATASET, X
ACCMETH=VSAM , X
BUFND=5, X
BUFNI=4, X
DATASET=CMFSERV, X
DISP=SHR, X
DSNAME=SAGLIB.NCIDB2.SERVER, X
FILSTAT=(ENABLED,CLOSED), X
JID=NO, X
LOG=NO, X
LSRPOOL=NONE, 1-8 ONLY FOR XA; NONE X
RECFORM=(FIXED,BLOCKED), X
RSL=PUBLIC, X
SERVREQ=(ADD,UPDATE,DELETE,BROWSE), X
STRNO=4
Restarting CICS is required, because of the additional FCT entry above.
Under Com-plete, the installation procedure of NDB continues with the adaptation of your Com-plete environment.
Ensure that the changes required for DB2 have been applied to your Com-plete environment (see the relevant Com-plete documentation).
This section describes how to install Natural for DB2 in an IMS TM environment.
Ensure that your Natural/IMS roll buffer is large enough to contain the
DB2SIZE; if you use the Natural Tools for DB2, an additional
storage of 8 KB is required.
Below is information on:
Job I055 / Steps 1631, 1632, 1633, 1634 for IMS MPP conversational, IMS BMP Natural, IMS MPP non-conversational, OBMP Natural
If desired, change library names and plan names to meet site requirements.
If the name (or any ALIAS) of your environment-dependent
Natural nucleus does not match the name of your DB2 plan, you must use an
Resource Translation Table (RTT).
Below is information on:
Add an additional DSNMAPN macro to your RTT as follows
(for any other parameters, refer to the relevant DB2 literature by IBM):
DSNMAPN macro:
DSNMAPN APN=load-module,PLAN=plan-name
The load-module represents
the environment-dependent Natural nucleus (that is, the IMS application
program) and the plan-name is the same
as the one used in the BIND step.
Be aware that database loops cannot be continued across terminal I/Os without using the File Server.
If you want to use the Natural File Server (VSAM), perform the following additional steps:
Job I008, Step 1600
Specify the size and the name of the VSAM RRDS that is to be used as the file server (see also Installing the File Server in Natural File Server for DB2).
Job I075, Step 1600
Specify the five input parameters required to format the file server dataset (see also Installing the File Server in Natural File Server for DB2).
Include the DD statement CMFSERV to
define the file server dataset.
Increase the REGION parameter if necessary.
Restart your MPP region, because of the additional DD
statement.
This section describes how to install Natural for DB2 in a TSO environment:
If you want to use the Natural File Server (VSAM), perform the following additional steps:
Set the NDBFSRV parameter in the
NATTSO macro to YES and reassemble and relink your
Natural/TSO interface NATTSO.
Job I008, Step 1620
Specify the size and the name of the VSAM RRDS that is to be used as the file server (see also Installing the File Server in Natural File Server for DB2).
Job I075, Step 1620
Specify the five input parameters required to format the file server dataset (see also Installing the File Server in Natural File Server for DB2).
To test the TSO installation of NDB under CAF, perform the following steps:
Job I070, Step 240C
Change the library and program names in the CLIST
NDBCAF to meet site requirements. If you do not use the file
server, remove the ALLOC and FREE statements for
CMFSERV.
Invoke Natural by executing the CLIST created in the previous step. Ensure that DB2 tables can be accessed and that plan switching can be performed.
Before the first SQL call you must call NATPLAN to
explicitly allocate the plan. The plan name must be the same as the name used
in Step 4 - Create NDB plan of
NDB Common Installation
Steps. NATPLAN can be edited to specify the
appropriate DB2 subsystem ID.
To test the TSO installation of NDB under DSN, perform the following steps:
Job I070, Step 240B
Change the subsystem ID as well as the library, plan and program
names in the CLIST NDBTSO to meet site requirements. If you do not
use the file server, remove the ALLOC and FREE
statements for CMFSERV.
Invoke Natural by executing the CLIST created in the previous step.
Ensure that DB2 tables can be accessed. The plan name must be the same as the
name used in the BIND step. For an explanation of the
DSN and RUN commands,
refer to the relevant IBM literature for DB2 TSO and batch users.
This section provides example batch jobs and online methods for verifying the installation of the Natural interface to DB2:
NDBBATCA contains sample JCL to test NDB in batch mode by
using the Call Attachment Facility (CAF) interface.
Modify the sample JCL to meet site requirements.
Before the first SQL call you must call NATPLAN to
explicitly allocate the plan. The plan name must be the same as the name used
in Step 4 - Create NDB plan of
NDB Common Installation
Steps. NATPLAN can be edited to specify the
appropriate DB2 subsystem ID.
NDBBATTB contains sample JCL to test NDB in batch mode by
using the DSN command processor. Modify the sample JCL to meet site
requirements.
The plan name must be the same as the name used in Step 4 -
Create NDB plan of NDB Common Installation
Steps. For an explanation of the
DSN and RUN commands,
refer to the relevant IBM literature for DB2 TSO and batch users.
NDBMTV01 contains a sample JCL to execute Natural by using
the DB2 DL/I batch support.
Modify the sample JCL to meet site requirements.
The plan name must be the same as the name used in Step 4 - Create NDB plan of NDB Common Installation Steps.
To verify the installation of the Natural interface to DB2 online, you
can use either SQL Services or DEM2 example programs:
To verify and check the NDB installation by using the SQL Services
of the Natural SYSDDM utility
Invoke Natural.
Invoke SYSDDM.
On the SYSDDM main menu enter Function Code
B to invoke the SQL Services function.
Enter Function Code S to select all DB2 tables.
The communication between Natural and DB2 works if all existing DB2 tables are displayed.
For one of the tables, generate a Natural DDM as described in the section Generate DDM from an SQL Table in DDM Generation.
After you have generated a DDM, access the corresponding DB2 table with a simple Natural program:
Example:
FIND view-name WITH field = value DISPLAY field LOOP END
If you receive the message NAT3700, enter the Natural system command
SQLERR to display the corresponding SQL return code.
You can find the description of the SQLERR command
in the section Natural
System Commands for DB2 in Natural Tools for
DB2.
To verify and test your installation you can also use the example
programs DEM2* in the Natural system library SYSDB2
provided on the installation tape.
Using these example programs, you can create a DB2 table by using
DEM2CREA and create the corresponding DDM using the Natural
SYSDDM utility.
You can then store data in the created table by using
DEM2STOR and retrieve data from the table by using
DEM2FIND or DEM2SEL.
You can also drop the table by using program
DEM2DROP.
This section covers the following topics:
To set the Natural profile parameter
Add the following Natural profile parameter to your
NATPARM module:
DB2SIZE=nn
The DB2SIZE parameter can also be specified
dynamically. It indicates the size of the DB2 buffer area, which must be set to
at least 6 KB.
The setting of DB2SIZE also depends on
whether you use the file server or not. If the file server is not
used, the setting can be calculated according to the following formula:
((1064 + n1 * 40 + n2 * 120) + 1023) / 1024 KB
If the file server is used, the setting can be calculated according to the following formula:
((1160 + n1 * 40 + n2 * 160 + n3 * 8) + 1023) / 1024 KB
The variables n1,
n2 and
n3 correspond to:
n1
|
the number of statements for dynamic access as specified as the second parameter in Job I055, Step 1600; |
n2
|
the maximum number of nested database loops as specified
with the MAXLOOP parameter in
NDBPARM;
|
n3
|
the maximum number of file server blocks to be allocated per
user specified as the fifth parameter in Job I075, Step 1620 or the
EBPMAX parameter of NDBPARM, if you decided
to use the Software AG Editor buffer pool as file server.
|
Important:
Ensure that you have also added the Natural parameters
required for the Software AG Editor; see the relevant installation description
in the section Installing the
Software AG Editor of the Natural
Installation documentation.
As DB2SIZE applies to Natural for DB2 and
Natural for SQL/DS, it must be set to the maximum value if you run more than
one of these environments.
Add an NTDB entry specifying the list of
logical database numbers that relate to DB2 tables. All Natural DDMs that refer
to a DB2 table must be cataloged with a DBID from this list. DBIDs can be any
number from 1 to 254; a maximum of 254 entries can be specified. For most user
environments, one entry is sufficient.
Important:
Ensure that all DB2 DDMs used when cataloging a given
program have a valid DB2 DBID. Also ensure that the DBIDs selected in the
NTDB macro for DB2 do not conflict with DBIDs selected
for other database systems.
The DBID for SQL/DS used when cataloging a Natural program does not
have to be in the NTDB list of DBIDs used when executing
this program. Therefore, when executing existing Natural programs, DBID 250 is
not mandatory. Two sample NTDB macros follow:
NTDB DB2,250
NTDB DB2,(200,250,251)
During execution of an SQL statement, storage is allocated dynamically to build the SQLDA for passing the host variables to DB2.
In previous Natural for DB2 versions, this storage was always obtained
from the TP monitor or operating system. For performance reasons, it is now
first attempted to meet the storage requirements by free space in the Natural
for DB2 buffer (DB2SIZE). Only if there is not enough
space available in this buffer, the TP monitor or operating system is
invoked.
To take advantage of this performance enhancement, you must specify
your DB2SIZE larger than calculated according to the
formula;
see Natural
Profile Parameter Settings.
Depending on the SQL execution mode and on the usage of the Natural file server, the additional storage requirements (in bytes) can be calculated as follows:
With sending fields:
80 + n * 56
With sending fields including LOB columns:
80 + 2 * n * 56
where n is the number of sending fields in an SQL statement.
The storage is freed immediately after the execution of the SQL statement.
With receiving fields (that is, with variables of the
INTO list of a SELECT statement):
80 + n * 56 + 24 + n * 2
With receiving fields including LOB columns:
80 + 2 * n * 56 + 24 + n * 2
where n is the number of receiving fields in an SQL statement.
The storage remains allocated until the loop is terminated.
With sending fields:
80 + n * 24
With sending fields including LOB columns:
80 + 2 * n * 56
where n is the number of sending fields in an SQL statement.
The storage is freed immediately after the execution of the SQL statement.
With receiving fields (that is, with variables of the
INTO list of a SELECT statement):
80 + n * 24 + 24 + n * 2
With receiving fields including LOB columns:
80 + 2 * n * 56 + 24 + n * 2
where n is the number of receiving fields in an SQL statement.
The storage remains allocated until the loop is terminated.
When using the file server, additional storage is required for each
database loop that contains positioned UPDATE and/or
DELETE
statements.
For each of such loops, a buffer is allocated to save the contents of
all receiving fields contained in the INTO list. Therefore, the
size of this buffer corresponds to the total length of all receiving
fields:
20 + 4 + sum (length (v1), ..., length (vn))
where v1 ... vn refers to the
variables contained in the INTO list.
The buffer remains allocated until the loop is terminated.
If you use the default value 10 for both variables
(n1 and
n2), the calculated
DB2SIZE will be 2208 bytes. However, if you specify a
DB2SIZE of 20 KB instead, the available space for dynamically
allocated storage will be 18272 bytes, which means enough space for up to
either 325 sending fields or 313 receiving fields.
Since space for receiving fields remains allocated until a database loop is terminated, the number of fields that can be used inside such a loop is reduced accordingly: for example, if you retrieve 200 fields, you can update about 110 fields inside the loop.
When using VARCHAR fields (that is, fields with either an
accompanying L@ field in the Natural view or an explicit
LINDICATOR clause), additional storage is allocated dynamically if
the L@ or LINDICATOR field is not specified directly
in front of the corresponding base field. Therefore, always specify these
fields in front of their base fields.
The source module NDBPARM is used in several Natural add-on
products. It contains parameter macros specific to an SQL environment:
These macros are described below.
The default values of the parameters contained in this macro can be modified to meet site-specific requirements (see the corresponding step of the NDB Installation Procedure, NDB Common Installation Steps). The values of the parameters cannot be dynamically overwritten.
Below is a description of all parameters contained in the
NDBPRM macro:
BTIGN |
CONVERS |
CONVRS2 |
DDFSERV |
DELIMID |
EBPFSRV |
EBPPRAL |
EBPSEC |
EBPMAX |
ETIGN |
FSERV |
MAXLOOP |
NNPSF |
PSCIGN |
REFRESH |
RETRYPO |
RWRDONL |
STATDYN
BTIGN ignores the error which results from a
BACKOUT TRANSACTION statement that
was issued too late for backing out the current transaction, because an
implicit Syncpoint has previously been issued by the TP monitor.
Possible Values:
| Value | Explanation |
|---|---|
ON |
The error after a late BACKOUT
TRANSACTION is ignored. This is the default value.
|
OFF |
The error after a late BACKOUT
TRANSACTION is not ignored.
|
This parameter is used to allow conversational mode in CICS environments where no NDB file server is used.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Conversational mode is allowed. This is the default value. |
OFF |
Conversational mode is not allowed. |
If this parameter is set to OFF and no Natural file
server is used, you cannot continue database loops across terminal I/Os; if so,
the DB2 SQL codes -501, 504, 507, 514, or 518 may occur.
If you are using the SYSDDM SQL
Services in a CICS environment without NDB file server,
you must specify CONVERS=ON, otherwise you get the errors
mentioned above.
This parameter is used to allow conversational mode 2 in CICS environments.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Conversational mode 2 is allowed. |
OFF |
Conversational mode 2 is not allowed. This is the default value. |
This parameter is used to control conversational mode 2 in CICS
environments. Conversational mode 2 means that update transactions are spawned
across terminal I/Os until either an explicit COMMIT or explicit
ROLLBACK has been issued (Caution: DB2 and CICS resources are kept
across terminal I/Os!). This means CONVRS2=ON has the same effect
as the Natural parameter PSEUDO=OFF, except that the
conversational mode is entered after an DB2 update statement
(UPDATE, DELETE, INSERT) and left again
after a COMMIT or ROLLBACK, while
PSEUDO=OFF causes conversational mode for the total Natural
session.
See also CALLNAT subprogram NDBCONV, which
allows setting or resetting conversational mode 2 dynamically.
This parameter specifies a DD name for the NDB
file server module other than CMFSERV.
Possible Values:
| Value | Explanation |
|---|---|
| DD-name | Any valid DD name. There is no default value. |
This parameter determines the escape character to be used for generating delimited SQL identifiers for the column names and table names in SQL statements. A delimited identifier is a sequence of one or more characters enclosed in escape characters. You must specify a delimited identifier if you use SQL-reserved words for column names and table names, as demonstrated in the Example of DELIMID below.
Possible Values:
| Value | Explanation |
|---|---|
" |
Double quotation mark |
' |
Single quotation mark |
| None | No value: Delimited identifiers are not enabled. This is the default value. |
To enable generation of delimited identifiers,
DELIMID must be set to double quotation mark
("") or single quotation mark (').
The escape character specified for DELIMID and
the SQL STRING DELIMITER are mutually exclusive. This
implies that the mark (double or single quotation) used to enclose alphanumeric
strings in SQL statements must be different from the value specified for
DELIMID. If you enable delimited identifiers, ensure
that the value specified for DELIMID also complies with
the SQL STRING DELIMITER value of your DB2 installation.
See also the RWRDONL parameter to
determine which delimited identifiers are generated in the SQL string.
In the following example, a double quotation mark ("") has been specified as the escape character for the delimited identifier:
Natural statement:
SELECT FUNCTION INTO #FUNCTION FROM XYZ-T1000
Generated SQL string:
SELECT "FUNCTION" FROM XYZ.T1000
This parameter is used to determine whether the Natural file server uses the Software AG Editor buffer pool as storage medium or not.
Possible Values:
| Value | Explanation |
|---|---|
ON |
The parameter must be set to ON,
if the file server is to be used in a Sysplex environment. This is the default
value.
|
OFF |
The file server uses the VSAM file as in previous versions. |
This parameter specifies the number of blocks to be allocated primarily to each user of the Natural file server, if the Software AG Editor buffer pool is used as storage medium.
Possible Values:
| Value | Explanation |
|---|---|
0 - 32676 |
Number of blocks to be allocated primarily. |
20 |
This is the default value. |
If the EBPFSRV parameter is set
to OFF, EBPPRAL is not used at runtime.
This parameter specifies the number of blocks to be allocated secondarily to each user of the Natural file server if the Software AG Editor buffer pool is used as storage medium. The secondary allocation is used to allocate buffer pool blocks to the user if the primary allocation amount is already exhausted.
Possible Values:
| Value | Explanation |
|---|---|
0 - 32676 |
Number of blocks to be allocated secondarily. |
10 |
This is the default value. |
If the EBPFSRV parameter is set
to OFF, EBPSEC is not used at runtime.
This parameter specifies the maximum number of blocks to be allocated to each user of the Natural file server if the Software AG Editor buffer pool is used as storage medium. This parameter serves as upper limit for the allocation of buffer pool blocks to a single user.
Possible Values:
| Value | Explanation |
|---|---|
0 - 32676 |
Maximum number of blocks to be allocated. |
100 |
This is the default value. |
If the EBPFSRV parameter is set
to OFF, EBPMAX is not used at runtime.
This parameter is relevant in IMS TM MPP and message-oriented BMP environments only.
It is used to handle END TRANSACTION statements in a
message-driven IMS region (MPP or message-oriented BMP).
In such a region, an END TRANSACTION cannot be executed
by the Natural/IMS interface and is therefore ignored without any notification.
In such situations, the ETIGN parameter can be used to
issue an error message instead.
Possible Values:
| Value | Explanation |
|---|---|
ON |
The END TRANSACTION error is
ignored and processing is continued. This is the default value.
|
OFF |
The END TRANSACTION error is
not ignored.
|
This parameter determines whether the Natural file server is to be used and whether it can be disabled in the case of an initialization error.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Natural file server is to be used. |
OFF |
Natural file server is not to be used. This is the default value. |
DIS |
Natural file server is to be used but is to be disabled if it cannot be initialized. |
If FSERV is set to ON and the file
server is not operational, the initialization of the Natural SQL Gateway is
terminated with a corresponding Natural error message. The Natural SQL
Gatewayis disabled and any SQL call is rejected with a corresponding error
message.
This parameter specifies the maximum possible number of nested database loops accessing SQL databases.
Possible Values:
| Value | Explanation |
|---|---|
1 - 99 |
Maximum possible number of nested database loops. |
10 |
This is the default value. |
This parameter changes the sign character of positive Natural
variables which have format N, if they are filled from the SQL database system.
Usually these variables have the C as positive sign character. If
the parameter NNPSF is set to ON,
F is used as positive sign character.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Positive numbers put into Natural numeric
variables by the SQL database system get the sign F.
|
OFF |
Positive numbers put into Natural numeric variables by the SQL database system remain unchanged. This is the default value. |
This parameter influences the treatment of positive sqlcodes returned
from the SQL database system. If the parameter PSCIGN is
set to OFF, a NAT3700 error message is issued. If the parameter
PSCIGN is set to ON, positive sqlcodes are
treated as if they were zero, that is, no NAT3700 error message is issued.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Positive sqlcodes are treated as zero. |
OFF |
Positive sqlcodes cause a NAT3700 error message. This is the default value. |
This parameter is used to automatically set the DB2 server and package
set to the values that applied when the last transaction was executed. Server
and package set are refreshed by using the CONNECT TO
server-name and SET CURRENT PACKAGESET
= 'package-name' SQL statements of DB2.
Possible Values:
| Value | Explanation |
|---|---|
ON |
An automatic refresh is performed every time before a database transaction starts and if a server or package set has been specified. |
OFF |
No automatic refresh is performed. This is the default value. |
This parameter delimits the number of retries done by NDB in order to reposition a dynamic scrollable cursor in a pseudo-conversational environment (IMS MPP or CICS).
Possible Values:
| Value | Explanation |
|---|---|
0 - 2147483648 |
Number of retries done by NDB. |
10 |
This is the default value. |
This parameter applies only for dynamic scrollable cursors.
In pseudo-conversational environments, cursors are closed at terminal I/O. For dynamic scrollable cursors the current absolute position number and the current key column values are saved. After terminal I/O the dynamic scrollable cursor is opened again and positioned absolutely to the position of the saved absolute position. The contents of the key columns are compared with the saved values. If they match, processing continues with the next requested database operation.
If the contents of the key columns do not match the saved values, the
next rows are fetched and compared with the saved values until either the
values match or no row is found or the RETRYPO count is
exhausted. In the latter cases the cursor is repositioned to the saved position
and the prior rows are fetched and compared until either the values match or no
row is found or the RETRYPO count is exhausted. In the
latter cases a NAT3703 error message is issued. If a row is fetched whose key
columns matches the saved values, processing continues with the next database
instruction.
RETRYPO delimits the retries in each direction
(next or prior).
If RETRYPO is zero no repositioning takes
place.
This parameter determines which identifiers are generated as delimited
identifier in an SQL string. RWRDONL only takes effect
if the setting of the DELIMID parameter allows delimited
identifiers.
Possible Values:
| Value | Explanation |
|---|---|
ON |
Only identifiers that are reserved words are
generated as delimited identifiers. The list of reserved words is contained in
the NDBPARM macro. This list has been merged from the lists of
reserved words for DB2 for z/OS, DB2 for VSE/VM, DB2 for LINUX, OS/2, Windows
and UNIX, and ISO/ANSI SQL99.
This is the default value. |
OFF |
All identifiers are generated as delimited identifiers. |
This parameter is used to allow dynamic execution of statically generated SQL statements if the static execution returns an error.
Possible Values:
| Value | Explanation |
|---|---|
NEVER |
Dynamic execution is never allowed. This is the default value. |
ALLWAYS |
Dynamic execution is always allowed after an error. |
SPECIAL |
Dynamic execution is allowed after special
errors only.
These special errors are:
|
The parameter macro NDBID determines the database type of an SQL DBID.
The NDBID macro is specified as follows:
Default Database Definition
The default database type is specified as follows. It applies to all
database IDs not explicitly specified by NDBID.
NDBID=database-type
Single Database Definition
A single database ID and its type is specified as follows:
NDBID=database-type,database-id
Multiple Database Definition
Multiple database IDs of the same database type can be specified together, enclosed in parentheses:
NDBID=(database-type,database-id1,database-id2,...)
| Possible Values | Explanation |
|---|---|
DB2 |
Databases are accessed via NDB. This is the default value. |
| Possible Values |
|---|
1-254 |
To be able to use the Natural Tools for DB2 (see the relevant section), consider the following requirements and recommendations:
In order to be independent of DB2 versions, the Natural Tools for DB2
Retrieval and Explain functions have been designed not to access the DB2
catalog tables directly, but to access identical tables qualified by the
creator name SYSSAG.
Thus, before you can use the Retrieval or Explain functions, you must
create these tables. The SYSSAG tables must have the same columns
as the DB2 catalog tables and they must be created as ALIAS,
VIEW, or TABLE.
To help you create these tables, sample SQL code is provided in the
member DEMSQL4 in the Natural system library SYSDB2.
By default, it creates an ALIAS
SYSSAG.xxx for the corresponding
SYSIBM table.
For some catalog tables no indexes are defined. For performance reasons, consider creating copies of these tables with appropriate indexes.
For the following tables it is recommended to work with copies of the catalog tables:
SYSCOLAUTH
SYSDBRM
SYSFOREIGNKEYS
SYSINDEXPART
SYSKEYS
SYSSTMT
SYSSYNONYMS
SYSTABLEPART
SYSVIEWS
The CREATE TABLE and CREATE INDEX statements
required are included as comments in the sample SQL member
DEMSQL4. In addition, DEMSQLUP includes sample SQL
code to update the data in the copies of the catalog tables.
For any other table, it is recommended that you create an
ALIAS or a VIEW that points to the corresponding
SYSIBM table.
Note:
The sample SQL members can be executed with the ISQL
part of SYSDB2. ISQL enables you to read SQL members
from the Natural system library SYSDB2. To save an SQL member in
any other library, you can use the command LIBRARY
MYLIB in the ISQL input screen to switch to another library and
then save the SQL member. You cannot save SQL members in the library
SYSDB2.
These functions access DB2 PLAN_TABLEs. To use these
functions, a PLAN_TABLE must exist for your SQLID. For the layout
of the PLAN_TABLE, see the relevant DB2 documentation by IBM of
the EXPLAIN command.
It is recommended that you create an index on the following columns of
the PLAN_TABLE:
APPLNAME
PROGNAME
COLLID
QUERYNO
TIMESTAMP
DESC
QBLOCKNO
PLANNO
MIXOPSEQ
A Natural for DB2 (NDB) server stub is an interface module needed to communicate between the DB2 database system and the Natural server. The server stub module determines, sets up and invokes a Natural server environment for executing Natural stored procedures and Natural user-defined functions.
As mentioned in the Installation Procedure, there are two types of
server stubs: the NDB start server stub (STR) and the NDB server stub (SRV).
Both stubs are generated from the NDBSTUB macro.
The NDB start server is used for setting up the Natural server
environments desired. The start server stub must be the main execution program
in the Stored Procedure Address Space (SPAS). After the start server stub has
established the Natural server environments, it passes control to the
appropriate DB2 program (DSNX9WLM for WLM SPAS and
DSNX9STP for DB2 SPAS). When SPAS terminates, the DB2 program
returns control to the start server stub. The start server stub stops the
Natural server environments and returns control to the operating system.
The NDB start server stub reads the names and parameters of the Natural
server to be started from the CMSRVIN dataset.
CMSRVIN must be specified with DDNAME CMSRVIN.
The CMSRVIN dataset is a sequential file that contains all
information required to start the desired Natural servers. For each server to
be started, one START entry must be provided. The parameters used
for the START entries are identical to the parameters that apply
to the NDBSTUB macro. Enclose the contents of each
START entry in brackets and delimit comments by the following
signs: /* and */.
START=(SERVER=WDB42SRV,NATURAL=NATBAT4R,CMPRMIN=CMPRMIN,
CMPRINT=CMPRINT,CMTRACE=CMTRACE,THREADSIZE=768,
THREADNUMBER=2,TRACE=ON)
START=(SERVER=WDB4SSRV,NATURAL=NATBAT4R,CMPRMIN=CMPRMIN,
CMPRINT=CMPRINT,CMTRACE=CMTRACE,THREADSIZE=768,
THREADNUMBER=2,TRACE=ON)
/* START=(SERVER=QE42SRV,NATURAL=NATBAT41,CMPRMIN=QAPARM4, */
/* CMPRINT=CMPRINT,CMTRACE=CMTRACE,THREADSIZE=700, */
/* THREADNUMBER=2,TRACE=OFF) */
If the start server dataset is missing or has not been assigned, the start server stub will start a Natural server environment with the parameters that derive from the parameters defined for the start server stub itself.
The NDB server stub is the link between DB2 and Natural stored
procedures or Natural user-defined functions (Natural UDFs). Specify the NDB
server stub as EXTERNAL NAME in the
SYSIBM.SYSROUTINES table row that refers to the Natural stored
procedure or Natural UDF. The server stub is started by DB2/WLM when the
Natural stored procedures or Natural UDFs are invoked. The NDB server stub
creates a Natural session in the Natural server environment and invokes the
Natural subprogram comprising the Natural stored procedure or the Natural
UDF.
A Natural session created for executing a Natural stored procedure terminates when the corresponding Natural subprogram ends and control returns to DB2 and to the calling client.
A Natural session created for executing a Natural UDF stays active for
multiple function invocations if the PARALLEL attribute is set to
D and the FINAL CALL attribute is set to
Y. The session invoked for a Natural UDF function is terminated by
the server stub if it detects a termination call.
The JCL procedure of the Stored Procedure Address Space (SPAS) must
specify the NDB start server stub as program in the EXEC
statement.
The NDB start server stub and the NDB server stub must reside in a library contained in the steplib concatenation of the JCL procedure of the SPAS.
//************************************************************* //* JCL FOR RUNNING THE WLM-ESTABLISHED STORED PROCEDURES //* ADDRESS SPACE //* RGN -- MVS REGION SIZE FOR THE ADDRESS SPACE. //* DB2SSN -- DB2 SUBSYSTEM NAME. //* NUMTCB -- NUMBER OF TCBS USED TO //* PROCESS END USER REQUESTS. //* APPLENV -- MVS WLM APPLICATION ENVIRONMENT //* SUPPORTED BY THIS JCL PROCEDURE. //* //************************************************************* //DB27ENV2 PROC RGN=0K,APPLENV=DB27ENV2,DB2SSN=DB27,NUMTCB=8 //IEFPROC EXEC PGM=WDB42STR,REGION=&RGN,TIME=NOLIMIT, /* NDB start server stub //*IEFPROC EXEC PGM=DSNX9WLM,REGION=&RGN,TIME=NOLIMIT, // PARM='&DB2SSN,&NUMTCB,&APPLENV' //STEPLIB DD DISP=SHR,DSN=DSN710.RUNLIB.LOAD // DD DISP=SHR,DSN=CEE.SCEERUN // DD DISP=SHR,DSN=DSN710.SDSNLOAD // DD DISP=SHR,DSN=NATURAL.V2.TEST.NUCLEUS /* Library containing stubs and Natural nucleus //CMPRMIN DD DISP=SHR,DSN=SAG.SYSF.SOURCE2(TDB31PRM) /* Dynamic Natural parameters. //CMSRVIN DD DISP=SHR,DSN=SAG.SYSF.SOURCE2(CMSRVIN) /* Servers to be started. //CEEDUMP DD SYSOUT=X //SYSOUT DD SYSOUT=X /* Traces records of the NDB server stub //RMTRACE DD SYSOUT=X //CMPRINT DD SYSOUT=X //SYSPRINT DD SYSOUT=X //SYSERROR DD SYSOUT=X //SYSUDUMP DD SYSOUT=X
The NDBSTUB macro is used to generate the NDB server stub
and NDB start server stub. You can parameterize NDBSTUB to create
different stubs.
Below are the parameters available with NDBSTUB:
CMPRINT |
CMPRMIN
| CMTRACE |
GTRACE |
GTRCID |
MAIN |
MODE |
NATURAL
| SERVER
| THREADNUMBER
| THREADSIZE |
TRACE |
WLM
CMPRINT specifies the DDNAME of the CMPRINT
dataset to which the primary report output is written. If an asterisk (*) is
specified, a unique DDNAME Pnnnnnnn is
built whenever a Natural stored procedure is invoked.
Possible Values:
| Value | Explanation |
|---|---|
ddname |
8 character DDNAME |
CMPRINT |
This is the default value. |
CMPRMIN specifies the DDNAME of the CMPRMIN
dataset during startup to read the input PROFILE parameter for
this server.
Possible Values:
| Value | Explanation |
|---|---|
ddname |
8 character DDNAME |
CMPRMIN |
This is the default value. |
CMTRACE specifies the DDNAME of the CMTRACE
dataset to which the primary report output is written. If an asterisk (*) is
specified, a unique DDNAME Pnnnnnnn is
built whenever a Natural stored procedure is invoked which makes it possible to
store each output separately.
Possible Values:
| Value | Explanation |
|---|---|
ddname |
8 character DDNAME |
CMTRACE |
This is the default value. |
GTRACE specifies whether or not the server stub executes
GTRACE macro calls for tracing purposes.
Possible Values:
| Value | Explanation |
|---|---|
ON |
The generated server stub executes GTRACE macros
in order to document its processing.
|
OFF |
The generated server stub does not execute GTRACE
macros during its processing cycle.
This is the default value. |
GTRCID specifies the event ID recorded with the
trace data created by the NDB server stub.
Possible Values:
| Value | Explanation |
|---|---|
event-id |
Decimal number from 0 to
1023 |
203 |
This is the default value. |
The value of MAIN is no longer evaluated. The NDB server
stubs check whether they are invoked as IBM LE (Language Environment) main
program or as IBM LE subprograms and react accordingly.
| Value | Explanation |
|---|---|
YES |
The generated server stub operates as IBM Language Environment
main program.
This is the default value. |
NO |
The generated server stub operates as IBM Language Environment sub program. |
MODE determines the operating mode of the NDB
server stub generated.
| Value | Explanation |
|---|---|
STR |
The generated NDB server stub operates as NDB start server stub that sets up the Natural server environment. |
SRV |
The generated NDB server stub operates as NDB server stub that
invokes the associated Natural stored procedure or Natural UDF.
This is the default value. |
NATURAL denotes the name of the server
front-end or Natural server load module which will be loaded by the NDB server
stub if the external CMSTART is not already resolved by the
linkage editor during creation of the server stub. The named load module has to
be present in any steplib of the stored procedure address space.
| Value | Explanation |
|---|---|
name |
Any valid load module name |
NATBATvr |
This is the default value. |
where vr is the current
product version number.
Server names suffixed with the three characters SRV
denote the names of the servers used by the server front-end in order to
identify the Natural server. These names must be unique within one address
space.
| Value | Explanation |
|---|---|
server-name |
Up to 5 characters |
NDBvr |
This is the default value. |
where vr is the current
product version number.
THREADNUMBER determines the number of Natural
threads used by the Natural server. This number limits the number of Natural
stored procedures and Natural UDFs concurrently active in the Natural
server.
Note:
The value of THREADNUMBER is no longer evaluated.
Instead, the NDB start server stub uses the NUMTCB
parameter of the SPAS JCL procedure as THREADNUMBER
value. For further details, see the relevant DB2 literature by IBM.
| Value | Explanation |
|---|---|
number |
Decimal number |
10 |
This is the default value. |
THREADSIZE determines the size of the Natural
threads to be used by the Natural server. The size is specified in units of
kilobytes.
| Value | Explanation |
|---|---|
threadsize |
Decimal number |
768 |
This is the default value. |
Determines whether the NDB server stub generated writes trace records
or not. The trace records are written to the dataset specified with
DDNAME SYSOUT.
| Value | Explanation |
|---|---|
YES |
Trace records are written. |
NO |
No trace records are written. This is the default value. |
WLM (Workload Manager) specifies where control
is passed to after the NDB start server stub has established the Natural server
environments requested.
This parameter is only evaluated if the
MODE
parameter is set to MODE=STR. Specify WLM=YES if the
NDB start server stub runs in an address space that has been established by
WLM.
| Value | Explanation |
|---|---|
YES |
The start server stub generated links to
DSNX9WLM, after setting up the Natural server
environments.
|
NO |
The start server stub generated links to |