Restoring Replication on Primary Databases

Follow the steps in this section if you have upgraded a source database, or primary database, in a Replication Server system, or the destination database is also a source for other Replication Servers.

1. If you zeroed out the locator point before the upgrade or you have already resumed replication on the earlier section, you don't need to execute this step.

   Otherwise, clear the replication locator by executing the following command in the RSSD database:Log in to the RSSD for each replicated primary, and for each replicated RSSD, and issue:

   1> use RSSD_name
   2> go

   1> rs_zeroltm dataserver, database
   2> go

   Warning!  

   You must stop Replication Server before executing rs_zeroltm

2. Log in to each replicated primary, and replicated RSSD and issue:

   1> use database
   2> go

   1> dbcc settrunc ('ltm', 'valid')
   2> go

3. Restart the Replication Server.
4. If the database is also used as an RSSD, resume the Replication Server connection to the RSSD by issuing the following command to the Replication Server specifying the same string that was specified during the 'hibernate_on' command:

   1> sysadmin hibernate_off, 'Replication Server'
   2> go

5. Log in to the Replication Server, and resume the Log Transfer connection for each replicated primary and each replicated RSSD:

   1> resume log transfer from server.database
   2> go

   If this is a replicated RSSD, you must log in to the replicate Replication Server.
6. If you are using Rep Agent, log in to the Adaptive Server and restart the Rep Agent:

   1> use database
   2> go
   1> sp_start_rep_agent database
   2> go

7. If you are using LTM, restart the LTM.