Support Content Notification - Support Portal

Configuring Harvest to Work with Oracle 10g

Product/Component

Notification Id

8029

Last Updated

28 February 2021

Initial Publication Date

24 May 2019

1.0 Introduction
1.1 Download Oracle 10g ODBC Drivers

2.0 For All Servers

3.0 For New Servers on Windows

4.0 For Existing Servers on Windows

5.0 For New and Existing Servers on UNIX and Linux

6.0 Reconfiguring CAIPT/ODBC on UNIX and Linux
6.1 DBMS-related Environment Variables

1.0 Introduction

This document describes how to install the CAI/PT ODBC driver and how to manually configure the Harvest server installation to connect to an Oracle 10g database.

1.1 Download Oracle 10g ODBC Drivers

The Oracle 10g ODBC drivers can be downloaded below:

Har521Oracle10g.zip

2.0 For All Servers

For all Windows, UNIX, and Linux servers, do the following:

Ensure that the Oracle user is granted DBA and create user privileges.
Harvest does not support heterogeneous Oracle services. Therefore, when running the Harvest database on a remote machine, ensure that the local and remote machines are running the same version of Oracle.

3.0 For New Servers on Windows

These instructions apply to new users of Harvest 5.2.1 who intend to create a Harvest database.

Note: If you are using an existing schema that has been upgraded to Oracle 10g, see the next section, For Existing Servers on Windows.

Important! Do not install the Harvest server until instructed to do so.

To install CAIPT/ODBC on new Harvest servers on Windows, follow these steps:

If you have not already done so, download the latest Har521Oracle10g.zip file from Computer Associates Customer Support at http://ca.com/supportconnect. This document (harvest521_oracle10g_setup.html) is included in the Har521Oracle10g.zip file.
Unzip the file Harvest Oracle 10g Configuration Utility.zip from Har521Oracle10g.zip to a temporary location.
Install the Oracle 10g Server or Client Administrator on the local machine. For instructions, see your Oracle documentation.
If you plan to continue using Harvest 5.2.1, run the Harvest Oracle 10g Configuration Utility that you unzipped in Step 2 on the machine on which you plan to install the Harvest 5.2.1 server. To do this, go to the temporary location where the Oracle 10g Configuration Utility.zip was extracted, and run the Ora10gPS.exe file in the harvest sourceOra10gPSRelease folder.
Install the Harvest server on the machine mentioned in the previous step. For instructions, see the chapter "Installing on Windows" in the CA Software Change Manager Install Guide.

Note: During the installation, when you are prompted to select the Oracle version currently installed, select Oracle 9i. The installation wizard does not list Oracle 10g explicitly, but selecting Oracle 9i and performing the other steps in this section configures Harvest to work properly with Oracle 10g.
After the Harvest 5.2.1 server installation has successfully completed, install the CAI/PT ODBC drivers. To do so, complete the remaining steps.
Unzip the odbc311win.zip file from Har521Oracle10g.zip to a temporary folder.
From the temporary folder, run setup.exe to install the new drivers.

The installation program informs you which drivers it does and does not install.
If you need to install the drivers again, do the following:

Click Start, Run on the Windows taskbar.

In the Open box, browse to the temporary folder to where the new ODBC drivers were unzipped and select the file called setup.exe. Click the Open button.

At the end of the line that appears in the Open text box, enter "reinstall=true". A sample entry is:
```
 c:tempodbcdriverssetup.exe reinstall=true 
```
Click OK.

The ODBC drivers are reinstalled.

4.0 For Existing Servers on Windows

These instructions apply to users who have already installed the Harvest 5.2.1 server and want to upgrade their database schema to Oracle 10g.

Important! Before performing these steps, ensure that your Oracle database has been upgraded to Release 10g and that the Harvest data has been migrated to Oracle 10g.

To install CAIPT/ODBC on existing Harvest servers on Windows, follow these steps:

If you have not already done so, download the latest Har521Oracle10g.zip file from Computer Associates Customer Support at http://ca.com/supportconnect.

This document (harvest521_oracle10g_setup.html) is included in the Har521Oracle10g.zip file.
Unzip the odbc311win.zip file from Har521Oracle10g.zip to a temporary folder.
From the temporary folder, run setup.exe to install the new drivers.
When prompted, select the existing folder where the ODBC drivers are located.

The installation program informs you which drivers it does and does not install.
If you need to install the drivers again, do the following:

Click Start, Run on the Windows taskbar.

In the Open box, browse to the temporary folder to where the new ODBC drivers were unzipped and select the file called setup.exe. Click the Open button.

At the end of the line that appears in the Open text box, enter "reinstall=true". A sample entry is:
```
 c:tempodbcdriverssetup.exe reinstall=true 
```
Click OK.

The ODBC drivers are reinstalled.
Update the Harvest server ODBC data source setting, as follows:
Open Settings, Control Panel, Administrative Tools, Data Sources (ODBC).
On the ODBC Data Source Administrator dialog, select the System DSN tab, and locate the data source called harvest.

Select the Harvest data source and click Remove. Click Yes to confirm the removal.

To create a data source, click Add.

From the Create New Data source driver selection box, select the CA/PT Oracle 8 driver, and click Finish.
In the ODBC Oracle Driver Setup dialog, enter harvest as a data source name.

Note: You must use harvest.
In the ODBC Oracle Driver Setup dialog, for a local database, leave the server name blank; for a remote database, enter the service name that points to the Oracle instance that contains the Harvest database.

Click the Advanced tab and ensure that Enable Scrollable Cursors and Application Using Threads are checked.
Click OK to close the ODBC Driver setup window.
Click OK again to close the ODBC Administrator.

5.0 For New and Existing Servers on UNIX and Linux

To install CAIPT/ODBC on new and existing Harvest servers on UNIX and Linux, follow these steps:

If you have not already installed the Harvest server, install it now. For instructions, see the chapter "Installing on UNIX and Linux" in the CA Software Change Manager Install Guide.

Note: During the installation, when you are prompted to select the Oracle version currently installed, select Oracle 9i. The installation wizard does not list Oracle 10g explicitly, but selecting Oracle 9i and performing the other steps in this section configures Harvest to work properly with Oracle 10g.
If you have not already done so, download the latest Har521Oracle10g.zip file from Computer Associates Customer Support at http://ca.com/supportconnect. This file contains all the tar files for UNIX and Linux.

This document (harvest521_oracle10g_setup.html) is included in the Har521Oracle10g.zip file.
The downloaded zip file (Har521Oracle10g.zip) contains the odbc311_UnixInstallScripts.tar file, which has the common UNIX-Linux installation scripts to be used with the odbc311platform.tar.Z file for the specific platform.

Extract the odbc311_UnixUnstallScripts.tar and the odbc311platform.tar.Z file for your platform to a temporary directory.

The following platform-specific files are included:
odbc311aix4.tar.Z
odbc311hpu11.tar.Z
odbc311sol2.tar.Z
odbc311linux.tar.Z
odbc311hpitan.tar.Z
Copy the odbc311_UnixInstallScripts.tar file and your platform-specific tar file to the same temporary location on the UNIX or Linux machine where the Harvest server is located. You can optionally use FTP or another utility to copy the files.

Extract the contents of both tar files to the same location.
If you are upgrading CAI/PT ODBC, stop all processes using the CAI/PT ODBC libraries. Otherwise, some of those libraries will be locked and therefore will not be updated. If you have problems overlaying files from the downloaded compressed tarball, you may need to revert to single-user mode.
If you have defined ODBC data sources that you want to save and update for use with the latest ODBC drivers, do one of the following:
Set the global environment variable ODBCINI to point to the complete path name of the data sources (export the variable for Bourne/Korn/Bash shells; use setenv with C Shell). When you upgrade from an existing release of CAI/PT ODBC to a later release, the upgrade process refers to $ODBCINI to find the ODBC data source definition file. (The ODBCINI environment variable selects the ODBC data source file used in all CAI/PT ODBC operations). Typically, the ODBC data source files are named odbc.ini.
Copy the data source defining file as the file odbc.ini in the directory where CAI/PT ODBC will be installed (where $ODBC_HOME points for an upgrade).
If you already have third-party driver definitions (definitions not supplied by Computer Associates) in an odbcinst.ini file that you want to save and have converted to add the new drivers, copy that odbcinst.ini file into the current directory.

Your current active odbcinst.ini file most likely is the expansion of your current $ODBCINST environment variable. If that variable is not defined, search the directory indicated by your current $ODBC_HOME variable. To display the directory name, enter the following command:
```
 echo $ODBC_HOME; ls -F $ODBC_HOME 
```
Run the shell script INSTALL.SH to:
Install the latest CAI/PT ODBC files.
Convert any existing data source definitions to use the new drivers.
Upgrade your existing odbcinst.ini file (if any) for any new CAI/PT ODBC driver definitions.
Configure the ODBC environment setup shell scripts, odbcenv.sh, and odbcenv.csh.

Requires you to enter the target CAI/PT ODBC home (installation) directory when $ODBC_HOME does not exist or does not point to a valid CAI/PT ODBC installation. This directory can be an absolute or relative path. For example, the following relative path is valid:
```
 . #[a single period] 
```
When a valid installation exists, checks the currently installed ODBC version (with $ODBC_HOME) and compares it to the upgraded version.
If the upgrade is newer or of a different type, asks you if you want to upgrade. For example, the shell script detects directories with different names in each release, such as Release Libraries versus Debugging Libraries.
At your command, terminates the installation. To force a reinstallation of the same ODBC release, enter the following commands:
```
 INSTALL.SH force_install=true #[on UNIX Systems]
 bash INSTALL.SH force_install=true #[on Linux Systems] 
```
Checks and publishes all required DBMS-related environment variables for each DBMS supported by this release of CAI/PT ODBC. The shell script checks for the presence of required files and directories when possible but does not check the contents of any files.
Allows you to change or create any of these variables. For more information about these variables, see DBMS-related Environment Variables in this document.
Configures the new CAI/PT ODBC installation. For example, the shell script establishes the DBMS-related environment variables inside the environment setup scripts odbcenv.sh and odbcenv.csh.
See DBMS-related Environment Variables in this document for important information about how odbcenv.sh and odbcenv.csh work.

Execute the shell script odbcenv.sh or odbcenv.csh on your current shell to export the ODBC and DBMS environment variables set up during the installation.

Important! Every time you log in, you must rerun the shell script. Ensure that you inform all users of this requirement.

Examples:
. odbcenv.sh (for Bourne/Korn/Bash shells)
source odbcenv.csh (for csh shells)
Run the Oracle configdsn utility to create and configure ODBC data sources. The utility prompts you for the required configuration parameters for each DBMS and related ODBC driver. You choose the driver for each DSN. You can reconfigure or delete existing data sources with configdsn.

When prompted to provide an operational set of records separated by semicolons (;) enter the following:
```
 EnableScrollableCursors=1;ApplicationUsingThreads=1 
```
When you are satisfied that your environmental shell script is correct, add its execution to your .profile; ensure that you include the complete path name to the CAI/PT ODBC installation directory.

Examples:
where /opt/CA/caiptodbc is the CAI/PT ODBC installation directory. (It is the default.)

6.0 Reconfiguring CAIPT/ODBC on UNIX and Linux
After a successful installation, you can optionally reconfigure ODBC, including converting other data source definition files for new drivers, as follows.
. /opt/CA/caiptodbc/odbcenv.sh (for Bourne/Korn/Bash shells)
source /opt/CA/caiptodbc/odbcenv.csh (for C shell)
Set $ODBC_HOME to point to the CAI/PT ODBC installation directory. (The ODBC environment setup shell scripts, odbcenv.sh and odbcenv.csh, will do this for you.)
For each file you want converted, do the following:

Set the global environment variable ODBCINI to point to the full path for the target data source definition file. (The ODBCINI environment variable selects the ODBC data source file used in all CAI/PT ODBC operations.)

Note: If you want to reconfigure only the DBMS environment variables, you do not need to change the ODBCINI variable.

On UNIX systems, enter the following command. This command selects /bin/ksh to run successfully.
```
 $ODBC_HOME/CONFIG.SH . 
```
On Linux systems, enter the following command:
```
 bash $ODBC_HOME/CONFIG.SH. 
```
By default, the ODBC environment-setting shell scripts (odbcenv.csh and odbcenv.sh) that are configured by running CONFIG.SH do not change DBMS environment variables if they are already set. Therefore, after you have edited the CONFIG.SH file, source again, as follows:

For the Korn and Bash Shells (ksh and bash), the easiest method is to enter the following command:
```
 . odbcenv.sh -a 
```
For all shells (including the Korn and Bash shells), set the ODBC_SET_DBMS_VARS shell variable before executing the appropriate ODBC environment setting script.

For example, for the Bourne, Korn and Bash shells, enter the following command:
```
 ODBC_SET_DBMS_VARS=true . odbcenv.sh 
```
A sample command for the C Shell follows:
```
 set ODBC_SET_DBMS_VARS=true source odbcenv.csh 
```
Note: Neither the Bourne shell nor the C shell passes command line arguments when a script is sourced.

6.1 DBMS-related Environment Variables

This section describes the DBMS variables that you can set in your ODBC environment during the CAI/PT ODBC installation process. You can ignore the variables for any DBMS that you do not use. However, if any variables are not set for a DBMS that you do use, you must set them.

Variable Descriptions

ORACLE_HOME
Specifies the home directory of a working Oracle client installation. The CAI/PT ODBC installation process checks to ensure that the specified directory exists.
ORACLE_SID
Specifies the local Oracle service instance to be used when Oracle Net*8 networking layers are not used. It is required only to access a local instance without Oracle networking. Your selection is not checked.
SYBASE
Specifies the home directory of a working Sybase client installation. The CAI/PT ODBC installation process checks to ensure that the specified directory exists.
INFORMIXDIR
Specifies the home directory of a working Informix 7 client installation. The CAI/PT ODBC installation process checks to ensure that the specified directory exists.
INFORMIXSERVER
Specifies the logical name of the Informix 7 service to which you want to connect, as configured in the $INFORMIXDIR. INFORMIXSERVER is similar to ORACLE_SID; however, INFORMIXSERVER is required to specify either a local or remote service.
DB2INSTANCE
Specifies the DB2 client instance. This instance must also be a valid system user. The home directory for the $DB2INSTANCE system user is also the home directory for the DB2 client instance. An appropriate environment setup file at ~$DB2INSTANCE/sqllib is invoked in the CAI/PT ODBC environment setup shell scripts. The CAI/PT ODBC installation process checks to ensure that the following exist: the DB2INSTANCE user, the home directory (from /etc/passwd), and the setup files in that directory.
II_SYSTEM
Specifies the home directory of a working Ingres client installation. The CAI/PT ODBC installation process checks to ensure that the specified directory exists.
Criteria for Setting Variables

The odbcenv.sh and odbcenv.csh files set the previously listed environment variables only if the following requirements are met:
The DBMS variables were set during the CAI/PT ODBC installation or a later run of CONFIG.SH.
The environment variable is not already set before odbcenv.sh or odbcenv.csh is run.
For the Korn and Bash shell only, run odbcenv.sh with the -a option.
For all shells, set the ODBC_SET_DBMS_VARS shell variable to a non‑null value before running an ODBC environment setting script. This method is useful after reconfiguring the ODBC environment with the CONFIG.SH installation shell script.