Enterprise Replication slowness

Hi, I have 2 servers connected via Enterprise Replication. The master server OS : Solaris 10 IDS version: IDS 10.00.FC8X9 Slave server: OS: AIX 6 IDS version: IDS 11.50.FC5XI I have a table with 605 row size. It has a primary key and additional indexes. The table has 30 millions rows The table has the same structure on both servers. CDR_MEMQUEUE is set to 1G on both servers and the send and receive queues have 12G each. CDR_EVALTHREADS is set to 100,2 and there are 4 CPU VPs on each server, therefore 402 eval threads on each server. All columns are replicated and there is no where clause in the replicate definition meaning that the whole table is replicated. The table is created empty at the target server and I use "cdr sync" command to synchronize the data. Yesterday, I launched the "cdr sync" command and it took about 14 hours before starting writing data on the target table. Once it started writing data on the target server, it took only a few minutes to replicate the whole table. Has someone already experienced such issues? Why does it take that long (14 hours) to preprocess the data before sending it to the target? I have some other tables to replicate and some of them have around 200 millions rows. I am concerned about the duration. If someone has already run into the same issue, I would appreciate if he could share his story. Are there parameters that could be tuned to make it run faster? Are they known issues with these versions regarding Enterprise Replication? Thanks