Advanced Replication Performance problems

Hi Everyone,

I'm running Oracle 8.0.5.1.0 on a SUN Solaris 2.6 Unix platform.

I perform a delete of 6 records on a replicated table containing 6 million rows. The delete at the source replication site runs in seconds, whereas the propagated delete at the remote site takes over an hour! An explain plan of the original sql (at the source site) uses the appropriate index. However, the explain plan of the replicated sql (at the remote site) does a full table scan. The table structure and indexes at both sites are identical. How can I get the replicated sql to use the index at the remote site?

Note that the delete statement at the remote site originates from the $RP replication package procedure REP_DELETE. This statement is different from the original delete statement at the source site. Received on Thu Jun 10 1999 - 03:20:59 CDT