Re: ORA-1578...block corrupted...error is normal...a block...had a NOLOGGING...operation performed against
I understand and agree with all you have said.
I have begun departing from this recently after a recent nasty restore and recover that took us 3 weeks. Some of this delay was caused by successive failures of backup tapes. We had to go back 3 days as one tape each from the more recent backup sets failed. fortunately, the archivelogs were recovered. UNfortunately ALL online redos were lost, but that's another matter. One of these tapes failed on the second use. We completely restored from a backup set, then 24 hours later had to get some more files from the same backup set and one of the tapes had failed. rotten luck! This is an environment where all backups are verified and tapes are regularly.
This reminded me not to place too many eggs in the backup basket. If the most recent backup contained the first backup of the new segments created nologging and that tapeset fails then you are left in the same precarious situation.
The benefits to nologging are clear and inviting. i think we should remember the additional risks associated.
On 8/23/05, Jared Still <firstname.lastname@example.org> wrote:
On 8/22/05, Mercadante, Thomas F (LABOR) <
I was thinking about this whole topic over the weekend. It just affirms my feeling that using the NOLOGGING option needs to be used judiciously. And I question how much time we are saving here.
Quite a bit of time actually. Try timing a bulk load operation with and without logging.
If the total time to reload using NOLOGGING and a total backup is less than a reload with LOGGING and no backup, then don't use it.
You don't need a complete database backup, just the datafiles containing the nologging objects.
Certifiable Oracle DBA and Part Time Perl Evangelist