Fixing NOLOGGING changes on the standby database

It's possible to limit redo generation for specific operations on Oracle databases, which provide higher performance. These operations include bulk inserts, creation of tables as select operations, and index creations. When we work using the NOLOGGING clause, redo will not include all the changes to data on the related segments. This means if we perform a restore/recovery of the related datafile, or of the whole database after the NOLOGGING operations, it'll not be possible to recover the data created with the NOLOGGING option.

The same problem exists with Data Guard. When the NOLOGGING operation is executed in the primary database, Data Guard is not able to reflect all the data changes in the standby database. In this case, when we activate a standby database or open it in the read-only mode, we'll see the following error messages:

ORA-01578: ORACLE data block corrupted (file # 1, block # 2521)
ORA-01110: data file 1: '/u01/app/oracle2/datafile...