Hi,
I have a (non-RAC) database that is 11.2.0.3.9 backed up using RMAN, nocatalog. I notice that in the alert.log the message 'backup piece header validation failure for handle' followed by a backup piece name. There is usually a list of at least a dozen or more files. The timing occurs with the RMAN backup, with no error or message written as part of the backup. It seems that the backup piece listed is one that would have been deleted with the 'delete nonprompt obsolete'. The command set is as follows:
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '/backup_db/ptcrebsdb0p/rman/iprd/Wed/%F';
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 2 DAYS;
configure device type disk parallelism 4;
CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT '/backup/rman/Wed/ora_db_%U_%d_%T_%s.rman' MAXPIECESIZE 10G;
CONFIGURE CHANNEL 2 DEVICE TYPE DISK FORMAT '/backup/rman//Wed/ora_db_%U_%d_%T_%s.rman' MAXPIECESIZE 10G;
CONFIGURE CHANNEL 3 DEVICE TYPE DISK FORMAT '/backup//rman/Wed/ora_db_%U_%d_%T_%s.rman' MAXPIECESIZE 10G;
CONFIGURE CHANNEL 4 DEVICE TYPE DISK FORMAT '/backup/rman/Wed/ora_db_%U_%d_%T_%s.rman' MAXPIECESIZE 10G;
crosscheck backup;
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/backup_db/rman/Wed/snapcf_iprd_before.ctl';
run {
BACKUP as compressed backupset DATABASE include current CONTROLFILE plus ARCHIVELOG;
delete noprompt obsolete;
}
There is a nightly full backup as shown, with an hourly backup of just archive logs, but does have the 'delete ...obsolete' phrase as well. In the backup logs the only reference that I can find for the backup piece listed is during the backup and during the crosscheck when found available.
I assume this is normal??? It baffles me that I don't see this for any other of our databases that are being backed up with RMAN, with the same type command set.