Chapter 7. Known Issues and Limitations
PostgreSQL
Need to edit pg_hba.conf and postgresql.conf and allow tcp connections from local network address
SQL Server
uniqueidentifier and timestamp data types in SQL Server are not supported, as these data types need external functions for usage. DBReplicator will not be able to perform operations on tables having such columns.
Precision in Heterogeneous Databases
When dealing with heterogeneous databases, precision of data types should be considered. If the precision of publishing RDBMS is more than that of the subscribing RDBMS, then the following problems may occur:
You will not be able to subscribe to the publication
Inconsistent Data will result
Primary Key Constraint
While dealing with heterogeneous databases, if users are trying to subscribe a table having binary / varbinary data type column as the primary key from an external RDBMS ( SQL Server) to Daffodil DB, this operation is not permitted in DBReplicator as Daffodil DB does not allow creation of tables having binary / varbinary data type column as the primary key.
Identity Columns in SQL Server
While using DBReplicator with an SQL Server database as Publisher and if a table has an identity column as its primary key, in such cases users of DBReplicator will not be able to synchronize such tables but snapshot operations can be performed.
System Tables
Manually updating system tables may affect the functioning of the replicator. It may prevent synchronization of data, show an error message or give an undesired result.
Auto Increment Column
You can't replicate a table having just an autoincrement column as its primary key, unless you plan never to add records except in the publisher database. It will invariably cause conflicts. The best way to handle this scenario is to include a site identifier column of some sort in the primary key. Thus, two records might get the same number in their autoincrement field, but only two records from the same site would collide with each other.
