Sybase Replication Agent recognizes all the standard native datatypes used in the supported primary database servers. However, some database servers allow you to create “user-defined datatypes” in the primary database. When the Replication Agent encounters a non-standard datatype in the primary database, it does not recognize the data in that column, and it does not capture any data from that column when it records a transaction.
The primary key column(s) are used to generate a where clause in the transaction operations applied at the replicate database. If no data is sent for a primary key column, the SQL executed at the replicate database will fail because the where clause will be empty (or contain incomplete data).
To avoid this problem, do not identify a column that contains a non-standard datatype as a primary key column in a replication definition.
