Using proxy databases guarantees unique database names with the cluster, but it does not guarantee unique database IDs. The same database may have a different database ID before and after failover. Because the database IDs may change, system procedures are automatically recompiled after failover to make sure they do not use an incorrect or out-of-date database or object ID from sysprocedures.
During failover mode, Adaptive Server performs a domain check to make sure that, if there are system procedures with duplicate names in the two Adaptive Servers, the system procedure in the correct domain is run. This domain check is only performed in failover mode.
