Milestone 3: Prepare Db2 Repository

This milestone describes all the planning required for the setup and installation of the Code Pipeline Db2 repository for a new installation of Code Pipeline. For a description of all the objects in the Code Pipeline Db2 repository, see Repository-Tables.

If you are performing an upgrade instead, skip ahead to Milestone 4: Install Code Pipeline Using SMP/E.

Complete the following tasks to plan your Db2 Repository installation.

• • Task 3.1 Determine Basic Db2 subsystem information
  • Task 3.2 Plan an authorization strategy
  • Task 3.3 Plan Db2 Subsystem resource utilization
  • Task 3.4 Develop Db2 Code Pipeline naming standard
  • Task 3.5 Planning for other items

Task 3.1 Determine Basic Db2 subsystem information

Certain Db2 subsystem details are needed to execute the Code Pipeline Db2 repository installation jobs.

Subsystem name

What is the name of the Db2 subsystem where Code Pipeline will be installed?

Db2 Subsystem Name:

VCAT name

What is the catalog name to be used for the Code Pipeline storage group (if one is to be created)?

VCAT Name:

Batch SQL program

What is the name of the batch SQL plan (for example, DSNTEP2) that this program invokes?

Batch SQL Plan:

What is the name of the batch SQL program (for example, DSNTEP2) that is used at your site?

Batch SQL Program:

Library names

What are the data set names of the Db2 LOAD, EXIT, and RUNLIB libraries, respectively?

Db2 EXIT Library Name:

Db2 RUNLIB Library Name:

Task 3.2 Plan an authorization strategy

The authorization strategy needs to be understood before the repository is installed. Different privileges are required for different phases of the installation. See Installation-Authority-Used for details of which privileges are used to execute the installation jobs.

Installation AuthID

Under what AuthID will the installation of Code Pipeline be performed?

Repository AuthID

What AuthID will be assigned to the Code Pipeline repository owner?

Maintenance AuthID (Optional)

Set up a maintenance AuthID if users—who don’t already have authority to access the Code Pipeline repository database through existing system-level privileges—will need to:

• administer the Code Pipeline repository database directly for housekeeping, or
• access the Code Pipeline repository database for other maintenance reasons (for example, backup, reorganization, or bufferpool tuning).

Maintenance AuthID:

Task 3.3 Plan Db2 Subsystem resource utilization

Adding a new application to run on a Db2 subsystem creates an extra load on it. Resource limits and other maximum limits should be revised to ensure that the subsystem can easily handle the extra resources required by Code Pipeline.

Maximum threads

Determine the maximum number of concurrent threads that could be used by the Code Pipeline server to communicate with the Db2 repository. This is the number of Server Processes (SPROCESS + GPROCESS) that have been defined in the Code Pipeline Server Definition Initialization parameters, plus three internal processes:

Number of SPROCESSs:
Number of GPROCESSs: +
Internal processes: + 3

Total concurrent processes: = (Maximum)Make sure the values specified on the Db2 installation panels for MAX TSO CONNECT and MAX USERS are large enough to compensate for this number of concurrent threads.

Make sure the EDM pool is large enough to store Code Pipeline package and plan information. Make sure there is adequate virtual bufferpool space for Code Pipeline tablespaces and indexes. Make sure other Db2 pools are large enough to handle Code Pipeline. Check the Db2 catalog for the plan, package, and DBD sizes.

Other limits

Revise any other relevant subsystem level parameters to ensure they are appropriate.

Task 3.4 Develop Db2 Code Pipeline naming standard

If your organization does not have a naming standard to follow, one should be established for Code Pipeline Db2 objects. A sample naming standard is provided in Sample Db2 Naming Standards, and Your Db2 Naming Standards can be used to record your own naming standard.

Sample names

This sample naming standard provides the ability to run multiple version of Code Pipeline simultaneously by including version numbers in some names. The following placeholders are used in  Sample Db2 Naming Standards.

There is other information that is required for the installation of the Code Pipeline repository. Some items may need more planning than others, depending on the site configuration. This is also time to think of any other site-specific requirements that have to be addressed.

Bufferpools

Determine the bufferpool(s) that the Code Pipeline repository tablespaces and indexes will use. The repository installation process allows different bufferpools to be specified for tablespaces and indexes. If further placement is required, it should be done after the DDL has been generated. For approximate sizes of the Code Pipeline tablespaces and indexes, see Db2-Repository-Sizes. All Code Pipeline tablespaces have a page size of 4K.

Tablespace Bufferpool (#BPTBSP):

Index Bufferpool (#BPINDEX):

Repository size

Use the table in Db2-Repository-Sizes to determine whether the site is small, medium, large, or system-determined. Use the related #SIZE code (either SML, MED, LRG, or SYS).

Repository Size (#SIZE):

DASD requirements

Determine how the Code Pipeline repository Db2 data sets are going to be physically stored on DASD. Ensure there is enough DASD for Code Pipeline, taking into account the likely growth rate in the short and long terms. If SMS is used for data set placement, determine any changes needed to the ACS routines. If not, determine which volumes the Code Pipeline Db2 data sets will reside on.

Code Pipeline CM STC UserID

Determine the UserID or AuthID associated with the Code Pipeline CM started task. This value is used during the repository installation, because the UserID needs to be granted authority to execute the Code Pipeline plan.

Code Pipeline CM UserID (#CMUSER ID):