Sometimes it is useful to maintain multiple replicate copies of a table at a single site. A subscription specifies the database where Replication Server maintains the replicated data. You can create subscriptions for multiple copies of a table by creating the table in different databases at the same site and then creating subscriptions for each one.
If OLTP and decision-support clients are on the same LAN, one Replication Server can manage both the primary data and the replicate data. Figure 2-1 illustrates such an arrangement.
Figure 2-1: Single LAN decision support replicate
For best performance, the databases are usually maintained by different data servers. The subscriptions can request different subsets of the data to be maintained in each database, so the replicated copies do not have to be identical.
If you must have two copies of a table in the same database, you can use multiple replication definitions for a primary table. One replication definition could have publishers as the replicate table name, and the other publishers2. Multiple replication definitions are also useful if you want different replicates to receive different column subsets. See “Multiple replication definitions”.
Another way to update multiple tables in an Adaptive Server database is to use stored procedures. Code the multiple updates in the stored procedures and write Replication Server function strings to execute the stored procedures. You can also use replicated functions and stored procedures to update multiple tables.
