OEM takes Lot of time to show Error Queue

Question

I have configured Multi-Master Advanced Replication for 3 master sites (Using Oracle Enterprise Edition 9.2.0.1.0 on MS Windows2K). But I have observed a problem during this week that the size of SYSTEM tablespace has been growing rapidly. After further analysis it was observed that a major portion of the tablespace is covered by DEF$_AQCALL. It has nearly 21,763,626 rows on master definition site and increasing day by day while there are just less that 6650,400 rows in participating other master sites. Due to this larger size when I try to observe the Error Queue using OEM, It takes more than 45 minutes to just show the Errors transactions list. Please have a look on the following script, I have run:-
SQL> select count(*) from def$_aqcall;
COUNT(*)
21763626
SQL> select count(*) from deferror;
COUNT(*)
3
SQL> select count(*) from deftrandest;
COUNT(*)
1466
Transactions are being propagated normally. But If I have to see the Error Queue to check the error transactions for possible data conflicts, it takes a long time. I have moved this table to USERS tablespace and then move again to SYSTEM tablespace but alas its size doesn't vary. Can I truncate this table i.e. DEF$_AQCALL. As its size is going to be bigger and bigger due to increasing number of rows in it. what should I do ?
Thanks
Masood

Answer 1

You should read the documentation for multimaster replication because mulitimater replication between two databases is easy but the replication between more then two databases required complete knowledge and strong understanding of oracle server working as well as replication concept. If you not create replication between more then two database with proper considration then it will create a very big overhead on the server and situation which you already mention.
Moreover create Replication with API rather then GUI tools. it will includes more option to create replication environment.