Description

SharePlex provides the following scripts to automate the building of a configuration file to specify Oracle source objects.

• config.sql: configure all tables and optionally all sequences in the database.
• build_config.sql: configure multiple or all tables in a schema

Supported databases

Oracle

Use config.sql

The config.sql script enables you to build a configuration that lists all of the tables, and optionally all of the sequences, in all of the schemas of a database. This script saves time when establishing a high-availability replication strategy or other scenario where you want the entire database to be replicated to an identical secondary database.

Conditions for using config.sql

• Source and target table names must be the same.

• The script does not configure objects in the SYS, SYSTEM, and SharePlex schemas. These schemas cannot be replicated since they are system and/or instance-specific.
• The script does not support partitioned replication. You can use the copy config command to copy the configuration file that the script builds, then use the edit config command to add entries for tables that use partitioned replication. Activate the new configuration file, not the original one.
• You can use the edit config command to make any other changes as needed after the configuration is built.

To run config.sql

1. Change directories to the config sub-directory of the SharePlex variable-data directory. The config.sql script puts configurations in the current working directory, and SharePlex configurations must reside in the config sub-directory.

   cd /vardir/config

2. Log onto SQL*Plus as SYSTEM.

3. Run config.sql using the full path from the util sub-directory of the SharePlex product directory.

   @ /proddir/util/config.sql

Refer to the following table when following the prompts.

Prompt	What to enter
Target machine	The name of the target machine, for example SystemB.
Source database SID	The ORACLE_SID of the source (primary) Oracle instance, for example oraA. Do not include the o. keyword. The ORACLE_SID is case-sensitive.
Target database SID	The ORACLE_SID of the target (destination) Oracle instance, for example oraB. Do not include the o. keyword. The ORACLE_SID is case-sensitive.
Replicate sequences	Enter y to replicate sequences or n not to replicate sequences.
SharePlex oracle username	The name of the SharePlex user in the source database. This entry prevents the SharePlex schema from being replicated, which would cause replication problems. If a valid name is not provided, the script fails.

Note: The name assigned by SharePlex to the configuration is config.file. If you run the script again to create another configuration file, it overwrites the first file. To preserve the original file, rename it before you create the second one.

Next steps:

• If any tables or owners are case-sensitive, open the configuration file with the edit config command in sp_ctrl, then use the text editor to enclose case-sensitive table and owner names within double-quote marks, for example “scott”.“emp”. The script does not add the quote marks required by Oracle to enforce case-sensitivity.

  sp_ctrl> edit config filename

• To ensure that the configuration is in the correct location, issue the list config command. If the name of the configuration is not shown, it was created in the wrong directory. Find the file and move it to the config sub-directory of the variable-data directory.

  sp_ctrl> list config

Use build_config.sql

The build_config.sql script enables you to build a configuration that contains multiple (or all) tables in a schema. It is an interactive script that prompts for each component of the configuration step by step. Instead of entering the information for each object and the routing individually, you can use a wildcard to select certain tables at once, or you can select all of the tables in the schema.

Conditions for using build_config.sql

• Source and target table names must be the same.
• The script does not support sequences. Before you activate the configuration that the script builds, you can use the edit config command in sp_ctrl to add entries for sequences.
• The script does not support partitioned replication. You can use the copy config command to copy the configuration that the script builds, then use the edit config command to add entries for the tables that use partitioned replication. Activate the new configuration, not the original.
• The script does not configure objects in the SYS, SYSTEM, and SharePlex schemas. These schemas cannot be replicated since they are system and/or instance-specific.
• You can run build_config.sql for different schemas, then combine those configurations into one configuration by using a text editor. Make certain to eliminate all but one Datasource:o.SID line, which is the first non-commented line of the file. Do not move the file out of the config sub-directory.
• You can use the edit config command to make any other changes as needed after the configuration is built.

To run build_config.sql

1. Change directories to the config sub-directory of the SharePlex variable-data directory. The build_config.sql script puts configurations in the current working directory, and SharePlex configurations must reside in the config sub-directory.

   cd /vardir/config

2. Log onto SQL*Plus as SYSTEM.

3. Run build_config.sql using the full path from the util sub-directory of the SharePlex product directory.

   @ /proddir/util/build_config.sql

Refer to the following table when following the prompts.

Prompt	What to enter
Target machine	The name of the target machine, for example SystemB.
Source database SID	The ORACLE_SID of the source (primary) Oracle instance, for example oraA. Do not include the o. keyword. The ORACLE_SID is case-sensitive.
Target database SID	The ORACLE_SID of the target (destination) Oracle instance, for example oraB. Do not include the o. keyword. The ORACLE_SID is case-sensitive.
Owner of the source database tables	The owner of the source tables.
Owner of the target database tables	The owner of the target tables.
Table name to include (blank for all)	Do one of the following: Press Enter to accept the default, which selects all tables that belong to the source owner. Enter a wildcard (%) character and a string to select certain tables, for example %e_salary%. Enter an individual table name.
Name of the output file to create	A name for the configuration. The script gives the file a .lst suffix, for example Scott_config.lst.

Next steps:

• If any tables or owners are case-sensitive, open the configuration with the edit config command in sp_ctrl, then use the text editor to enclose case-sensitive table and owner names within double-quote marks, for example “scott”.“emp”. The script does not add the quote marks required by Oracle to enforce case-sensitivity.

  sp_ctrl> edit config filename

• To ensure that the configuration is in the correct location, issue the list config command. If the name of the configuration is not shown, it was created in the wrong directory. Find the file and move it to the config sub-directory of the variable-data directory.

  sp_ctrl> list config