OS : Solaris 5.10
DB ; Oracle 11g 11.2.0.4
Recently i got into a situation where i had to restore and recover the PROD database. (Panicked but was very confident that i have the RMAN backup and I will recover the database upto the failure point.)
After Restore and recover, on DB startup realized that there are multiple block corruption due to NOLOGGING Issue in 52 datafiles (dangerous situation, i can't recover the database out of this mess, no one can help, entire company existence on stake).
I checked the LOGGING=NO in tables, indexes, LOB, partitions wherever possible and nothing with this option, every where LOGGING=YES. (we don't use sql loader anywhere, rather we user externally identified tables). I followed the "How to identify all the Corrupted Objects in the Database with RMAN ( Doc ID 472231.1 )" Luckily in my case all the corrupted block in table segments were above the high water mark, to fix this i moved the tables in same tablespace and rebuild the indexes
questions:
1) do you place the database in force logging mode even if its standalone database ? (usually I do place the database in force logging mode when physical standby database is configured)
2) what could be the operation on database that still did the NOLOGGING when there is not object in database with NOLOGGING options (exception : MVIEW LOgs table RUPD$_%, which i ignored) and we are not using the sql Loader ?
Note:
Also to be provide more background : The outage happened on 20-Aug-16 .. Used the 18-aug-16 Level 0 full backup for restore, during the period there was no release deployment and there were no DDL operation executed on DB
Regards,
Anurag