Review the following storage requirements before you begin the product installation.
Review the following storage requirements for the
Database Management Solutions for Db2:
When installing the
Database Management Solutions for Db2, we recommend that you have 6,000 cylinders available with the following allocations:
- 200 cylinders for the downloaded files (approximately 371 unzip files)
- 5,600 cylinders for target, distribution, and SMP/E files
The cylinders total does not include the data sets that are created when you execute the post-installation tasks during product configuration. Temporary data is deleted after the installation process completes.
Complete the following requirements based on whether you are using
CSMor Pax-ESD to install the products:
- Increase the default region from within your Java JVM configuration by updating theCSMserver JCL member being used in the ddname STDENV:IJO="-Xms128m -Xmx768m"The name of theCSMserver is typically MSMTC.
- Increase the system setting for MAXCPUTIME to 7200.
- To display your current setting, specify:DISPLAY OMVS,OPTIONS
- To change MAXCPUTIME to the suggested limit temporarily, specify:SETOMVS MAXCPUTIME=7200
- To change MAXCPUTIME to no CPU limit, specify:SETOMVS MAXCPUTIME=86400
- To change MAXCPUTIME permanently, specify this value in BPXPRMxxin SYS1.PARMLIB, wherexxidentifies the active table suffix in your z/OS system.
- (Optional) Increase the system setting for MAXASSIZE using the following command:SETOMVS MAXASSIZE=2147483647To change MAXASSIZE permanently, update the BPXPRMxxmember.
- For Pax-ESD, ensure that you have sufficient free space in the UNIX System Services (USS) file system. The file system is used to hold the directory that the pax command and its contents create. You need approximately two times the pax file size in free space.
When there is not enough free space, you receive error message:
EDC5133I No space left on device.
Data Set Naming Conventions
During installation, the term
hlqdenotes the node-names concatenated to the low-level data set name qualifier library name as shown in the following examples:
- Single high-level node-name (PROD concatenated to a load library):PROD.CDBALOAD
- Multiple high-level node-names (PROD.RxxSP1.SMPE concatenated to a load library):PROD.RxxSP1.SMPE.CDBALOAD
When assigning these values, the high-level for the SMP/E TARGET data sets must be different from the high-level specified for the SMP/E TEMP data sets. For example:
PROD.RxxSP0.SMPE SMP/E TARGET high-level PROD.RxxSP0.SMPE.TEMP SMP/E TEMP high-level
Target and Distribution Libraries
The lowest data set level qualifier of the installed data sets is shown in the following list for the target libraries (CDBA prefix) and distribution libraries (ADBA prefix):
All target data sets begin with CDBA. All distribution libraries begin with ADBA. To create aliases relating old and new library names, see Customize Optional Global Settings.
- CDBACICS onlyLoad library for CICS
- CDBACLS0 and ADBACLS0CLISTs and REXX.
- CDBACOBI and ADBACOBICOBOL copybooks and source code.
- CDBADBRM and ADBADBRMDBRMs that are used by theDb2BIND process.
- CDBAEVNT and ADBAEVNTService desk members.
- CDBAFORM and ADBAFORMReport Facilityforms.
- CDBAGTF and ADBAGTFVariable length data.
- CDBAIMS onlyLoad library that contains load modules for IMS.
- CDBAJCL and ADBAJCLSample JCL.
- CDBALOAD onlyLoad library that contains executable code for theDatabase Management Solutions for Db2.
- CDBAMAC and ADBAMACAssembler macros.
- CDBAMAPS and ADBAMAPSSYSVIEW for Db2panels.
- CDBAMDL and ADBAMDLModels, similar to ISPF skeletons.
- ADBAMOD0 onlyObject modules.
- CDBAMSG0 and ADBAMSG0ISPF messages.
- CDBAPARM and ADBAPARMProduct options, inputs to batch jobs.
- CDBAPNL0 and ADBAPNL0ISPF panels.
- CDBAPROC and ADBAPROCJCL procedures.
- CDBAPXMP and ADBAPXMPXmanager parameters.
- CDBASAMP and ADBASAMPSamples that customers are expected to copy to their own data set for modification.
- CDBASIDE and ADBASIDESide decks.
- CDBASKL0 and ADBASKL0Bind skeletons that are required for each object with SQL.
- CDBASKL1 and ADBASKL1Information that is used byBind Analyzer.
- CDBASQL and ADBASQLSQL, DDL, and DML.
- CDBASRC and ADBASRCAssembler source.
- CDBATBL0 and ADBATBL0ISPF tables.
- CDBATREQ and ADBATREQSYSVIEW for Db2query language.
- CDBAVSMI and ADBAVSMIReport Facilitypanels.
- CDBAXML and ADBAXMLCSM.
- CDBAXMSG and ADBAXMSGXMSG contents.
- CDBAZFS (USS file)USS elements. To prevent GIM39302E errors when applying maintenance, ensure this file system is mounted correctly and that the CINSHFS and CSQ2HFS DDDEFs point to the correct USS directory so that shell script PDTAAMKD can be read.The CINSHFS, CSQ2HFS, and CSQ2HFSM DDDEFs are contained within CDBAZFS. These DDDEFs point to a USS path. The path must be the mount point of the zFS file system, plus the directory /CA. For example, if the zFS file system is mounted at /cai/dbmgmt, the DDDEF path must be /cai/dbmgmt/CA. CINSHFS is used with the base installation of theDatabase Management Solutions for Db2(FMID CINSK00). For example, the components that run out of the mounted zFS library. CSQ2HFS and CSQ2ZFSM are used with theCommon SQL Services for Db2(FMID SCQ2100) that are used byDatabase Management Solutions for Db2.
If required by an FMID, you could be prompted for other data set names. For example:
- CICS SDFHLOAD
- CETJSIDE (only used if FMID CEU9K00 is installed)
Various IBM data sets could also be required:
- CSSLIB (IBM linkage assist library)
- SCEEBND2 (64-bit processing)
- SCEECPP (Language Environment)
- SCEELIB (Language Environment)
- SCEELKED (Language Environment)
- SCEESPC (IBM System Programming C)
- SEZACMTX (IBM Socket Library)
If you update the external library versions in SMP/E, relink the
Broadcomproduct load libraries with the latest versions of these external modules. To detect and relink the load modules, execute the LINK LMODS CALLLIBS SMP/E command.
Recommended REGION Size for Executing the HDDL Command
The IEALIMIT or IEFUSI exit routine modules provide IBM-supplied default limits. If these defaults are not changed, the following region sizes are recommended to execute HDDL successfully:
Db2 Object Size
0 KB or 0 MB
Unlimited storage available (any object with any number of dependencies)
< 0 MB < REGION < 16 MB
Small (up to 100) and medium-sized (up to 1000) dependent objects.
< 16 MB < REGION < 32 MB
Large (more than 1000) dependent objects.
< 32 MB < REGION < 20478 MB
Any Db2 object with more dependencies.
Recommended MEMLIMIT Size for Generating DDL
In z/OS, the REGION and MEMLIMIT settings are used to limit the amount of virtual storage that is available for allocation. The overall application memory allocation above the 2-GB bar is limited by the MEMLIMIT specification.
For more information about MEMLIMIT, see the IBM z/OS documentation for 64-bit C/C++ and Java programming, installation exits, installation and tuning (JES2), and UNIX System Services planning.
To generate the DDL for any object type, Object Framework Services (OFS) retrieves all the
Db2objects that are required to generate all the clauses for that object type from the catalog table. After the objects are retrieved from the catalog table, they are stored inside OFS in an internal format. The storage size is determined by the number of
Db2objects that are required to generate the DDL.
- The DDL command generates the DDL for a single object type, so a small number of dependent objects are created inside OFS and less storage is used.
- The HDDL command generates hierarchical DDL, where a huge number of dependent objects are added into OFS and more storage space is used.
The following tables describe the minimum MEMLIMIT size that must be set to perform the DDL or HDDL operations:
In the following table, the size that is given for individual
Db2objects is the amount of storage that is needed during the DDL operation.
1.5 MB + 100 KB for every 50 volumes.
7 MB for every 500 partitions.
200 KB - 400 KB for every column in the table.
Native stored procedure
NSP with 20 parameters = 6 MB + SQL body size
1.9 MB + 5 KB for each parameter
The following table lists the required storage to perform the HDDL operation, which generates a hierarchical DDL. In this table:
- MEMLIMIT is the sum of storage that is needed for multiple objects.
- The row that is marked "Low" indicates the minimum required clauses to create the objects.
- The row that is marked "High" indicates all the possible clauses and the maximum number of supported columns or data types.
No. of Columns/Support Columns or Data Types
All objects except routines
All the objects from LOW columns of the DDL table except routines
All objects except routines
All the objects from MEDIUM columns of the DDL table except routines
All objects except routines
All the objects from HIGH columns of the DDL table except routines