Staging Tables

If your Replication Server replicate database connection has been configured to stage operations logged for a primary table, these operations are compiled where possible and written to staging tables on the replicate Sybase IQ database.

For each table to be staged, there are three staging tables, each corresponding to DELETE, INSERT, and UPDATE operations:

owner_table_name_DELETE_RS
owner_table_name_INSERT_RS
owner_table_name_UPDATE_RS

where owner and table_name are the owner and name of the corresponding primary database table. The names of these tables are generated by your EL project, and they cannot be changed.

Note: The Tables tab of your EL project displays only the insert staging table. However, the Table Creation window displays all three staging tables corresponding to a specified primary database table.

You must identify which primary database tables are to be staged in a Sybase IQ InfoPrimer EL project. You may also selectively exclude replicate tables from staging. For a table that has been excluded from staging, no corresponding staging tables need to be created, and data will be replicated from the primary table to a replicate table in the replicate Sybase IQ database.

If you configure a replicate database connection to stage tables but no staging tables exist in the replicate Sybase IQ database, the replicate database connection will be suspended. If a replication definition includes columns that are declared as identity columns, these will not be declared as identity columns in the corresponding staging tables.

Table Compilation

Compilation is not performed on noncompilable tables. Tables are considered noncompilable if they have RTL disabled, modified function strings, or minimal column replication enabled. Operations to noncompilable tables are captured in an ordered list and applied to the corresponding replicate table after compilation is complete.

Note: After Replication Server commits a staged operation, the transformation stored procedures truncate the corresponding staging tables. You should therefore not use the Replication Server rs_subcmp utility to validate staging tables.

Insert Staging Table Structure
Apart from changes and filtering applied by the corresponding replication definition, the insert staging table contains the same number of columns and the same column names as the primary table.
Delete Staging Table Structure
The delete staging table contains only the primary-key columns specified in the corresponding replication definition.
Update Staging Table Structure
The update staging table contains two columns for every primary-key column specified in the corresponding replication definition, one each for the column data before and after a change.

Parent topic: Using the Replication Server and Sybase IQ InfoPrimer Integration