I am having problem with Replication (8i).
The following steps have been carried out.
Relevant user TEST created in both Master Defn Site A and Master Site B
and granted execute on sys.dbms_defer to TEST, Tables exist with data.
1. Init.ora set for relevant entries
2. catrep ran in both the Master Defn site and Master Site (Lets call A
3. Globalname set to TRUE, dblink created etc, works fine
4. repadmin created, relevant grants given from sys , registered as
5. Grant any schema, lock any table blah,...
6.Created Replication group in A (Master Defn Site)
7.Added Object to Master Group
8.Generated Replication Support
Query from dba_repgroup shows group quisced. dba_repobject shows pkgs
generated and valid.
9.Added b.world as Master Database at A. Query in dba_repcatlog shows
Query from dba_repsites shows entries for A and B respectively.
From A, i do execute dbms_defer_sys.schedule_push to B.
From B, i do execute dbms_defer_sys.schedule_push to A.
From A, i do execute dbms_defer_sys.schedule_purge to B.
From B, i do execute dbms_defer_sys.schedule_purge to A.
Timing set as 10 seconds.
From the master Defn Site, i do RESUME_MASTER_ACTIVITY.
Do some inserts in A. Commit. Check for entries in B.
The damn thing doesn't work. It does not give any error. I see entries
deftran. I wait for 10 ,20, 30 minutes, but the propogation does not
Frustrated, i try DBMS_DEFER_SYS.DELETE_TRAN , PURGE etc but
the entry still remains. I check with DBA_JOBS, remove them by using
DBMS_JOB.REMOVE , still entry persists in deftran.
I drop repobject, repgroup , recreate a different group (after
OR whatever it requires), but it gets me nowhere.
Can anyone point out where the leak is?
Obviously I am missing something.
Thanks in Advance.
When Going Gets Tough, Tough Gets Going !!
It sounds like you did a good job setting up replication but perhaps the trouble is with either 1. the dB links and/or 2. the dbms jobs.
Originally posted by priyamvada
I see entries in deftran. I wait for 10 ,20, 30 minutes, but the propogation does not take place.
If your database links are not perfect then replication will not work!
1. Make certain SYSTEM has a PUBLIC dB link
2. Make certain there is a private dB link for repadmin and that you specified the CONNECT TO....IDENTIFIED BY. I used other types of connect methods and it failed me (connect user and current user do NOT work).
3. test the dB links with sqlplus
4. If links are working try and manual push!!!
5. What is the next_date in dba_jobs for repl jobs????
6. Make certain the jobs are owned by repadmin.
7. Why are you waiting 10, 20, 30 minutes if your interval is 10 seconds?????? If it dows not work in 10 seconds, it's not working!!! Try making your interval 60 or 30 seconds if the biz supports it.
My problems with replication began with dB links. When I was certian my links were okay, my problems moved onto the dbms_jobs but I eventually saw propogation with the manual push and purge.
I can tell that you are on your way to success. Repl is a pain in the a#%! but once it is working it basically works really well.
1. Query from dba_repgroup shows group quisced => this should be NORMAL.
2. Query in dba_repcatlog shows entries => It should return 0 rows. If there are records then check for the error message.
dbms_refresh.refresh('REPGROUP') from snapshot site
as user repadmin .And see what error it gives
In GOD I find everything
Drop me a mail. to firstname.lastname@example.org..with the following details..database name on both the sites..with the tnsnames.ora entries and job_queue_processes in the init.ora file settings...both sites...
Am work'n on multi-master Asynchronous repln...
Click Here to Expand Forum to Full Width