Kbase P143090: The restart of the Repl server on the source database fails with error (10716) when resourcing origi
| Autor |
Progress Software Corporation - Progress |
| Acesso |
Público |
| Publicação |
28/07/2009 |
|
Status: Verified
SYMPTOM(s):
The restart of the Repl server on the source database fails with error (10716) when resourcing original target database or when adding a second target database.
During an online backup a busy ai extent becomes locked (SEQ 11677) without a switch.
Restarting Repl server fails with error:(10716) Fathom Replication Agent agent# cannot be configured because the required AI area (area-number>-1, sequence number 11677 is not available.
area-number>-1 indicates that the ai file cannot be found.
(11696) The Agent agent# cannot be properly configured and is being terminated.
(10700) The Fathom Replication Agent agent# is being terminated.
(10504) Unexpected error -158 returned to function rpSRV_ServerLoop.
Unexpected error -158 indicates that the ai area is no longer active
(10505) The Fathom Replication Server is ending.
source/target.repl.recovery files were deleted.
Error getting rpserver and rpagent to synchronize after re-sourcing Target database.
FACT(s) (Environment):
OpenEdge 10.1B03 64-bit Service Pack
OpenEdge Replication 10.1C 32-bit
OpenEdge Replication 10.1B 64-bit
OpenEdge Replication 10.1B 32-bit
OpenEdge Replication 10.1C 64-bit
OpenEdge Replication 10.2A 32-bit
OpenEdge Replication 10.2A 64-bit
All Supported Operating Systems
CAUSE:
The source database was backed up with -REPLTargetCreation in order to resource the target or add a second target database. The existing Target, however, was still running and replicating. The subsequent ai files were unlocked and marked as empty. When the new Target was restored, the ai file (by sequence number) it expected was empty.
FIX:
When re-sourcing the Target or adding a second target database, make sure to stop the existing Target database. This will assure that the necessary ai files are locked and can not be emptied.
Another workaround would be to stop the Repl Server before taking the backup. This action will leave the ai files locked as the ai notes can not be replicated when the Repl server is not running. This will also allow the use of the target database during the time frame the repl server is down.
Commands for stopping and starting the replication server:
To stop the Repl-server issue the following command:
rfutil <dbname> -C terminate server
To restart the Repl-server issue the following command:
rfutil <dbname> -C restart server
If continuing for some time with the existing Source and Target, do not empty the ai files. This leaves the ai files as full.
If choosing either option, make sure you have enough free ai files for the time frame that replication of the ai notes will not be occurring.