Screw up. Database corruption happens. Properly designed backup procedures make recovery a straightforward process. They apparently had an ongoing problem and didn’t detect it before it had gotten into the backups. Regular integrity checks would have caught it before it got out of control.
Exactly. Even if you have a RAID or some other redundancy if the data itself is corrupt then it gets copied to the backup disc/system. If you do a restore and have a corrupted file/database it will get restored.
I have been bitten by this on several occasions and had to go back several generations of backups to find a non-corrupted file. Not fun at 2:00am in the morning...and those systems have to be up by 8:00am so production can resume..
Perhaps you’re right, but Cobol running on an IBM S/360 Mainframe is a beast that’s temperamental and hard to tame. /sarc.