Following errors keep appearing in source event log:
04/29/10 14:10 Notice: Oracle env - XP71:/u01/app/oracle/product/9.2.0 [sp_cnc(pdb)/18552]
04/29/10 14:12 Notice: 11009 - GetKeyValues failed: ORA-00904: "COLUMN_NAME": invalid identifier. [sp_ordr/23346]
04/29/10 14:12 Notice: 11009 - GetKeyValues failed: ORA-00904: "COLUMN_NAME": invalid identifier. [sp_ordr/23346]
04/29/10 14:14 Notice: 11009 - GetKeyValues failed: ORA-00904: "COLUMN_NAME": invalid identifier. [sp_ordr/23346]
The latest *ord* log (Reader log) has following entries:
10-04-29 13:00:29.096902 23346 1 ERROR: Failed to GetKeyValue with rc=904, objid=289754, actid=61, rowid=AABGvaAA3AACJLLABo
10-04-29 13:00:29.107585 23346 1 record with seqno=275187, offset=197458392
10-04-29 13:00:29.107623 23346 1 sqlStmt=select TO_CHAR("COLUMN_NAME", 'FM9.99999999999999999999999999999999999999999EEEE') from
10-04-29 13:00:29.107623 23346 1 "OWNER"."TABLENAME" T where rowid = :1
The name of the key column of replicating table was changed.
When the key column was changed, the object cache (Shareplex's representation of the structure of table) was not updated. This is causing the issues. To correct the problem:
1. Take the table out of replication by making a copy of the currently active config file, commenting out the problem table and activating the copy.
2. Then bring back the table back into replication by activating the original config file.
