Re: RMAN or not RMAN

AFAIK there is one fundamental difference between OS level datafile copy and RMAN copy. When using OS copy, you have to copy the whole datafile. RMAN copies only those datafile blocks that contain data and you can take incremental block level backups. This reduces time needed for taking backups, but you need a recovery catalog for storing backup information. There is one more feature -- RMAN verifies data block checksum on every recovery, so you can be sure that your backup is healthy. BTW, this checksum is stored in data block header, so when you recover your datafiles using RMAN, you won't get the *exact* copy of datafile as it was at momentum when backup was taken.

HTH,
Juhan

Scott Watson wrote:
>
> I was wondering if RMAN is really used or if DBA's are using shell scripts
> to perform backup and recovery.
>
> Are there any real performance improvements to using RMAN. The big issue
> that keeps coming up in the doc that I have read is it can be done in
> parallel, but I am not exactly sure what that means since I can always open
> another telnet session and recover other datafiles and therfore my recovery
> is being done in parallel??
>
> Any thoughts on the matter would be appreciated.
>
> Thanks,
> scott.
 

-- 
+-----------------+
|NB! Please remove|
|   .NICHT-SPAM   |
|  when replying  |
|    to get my    |
|     *REAL*      |
| e-mail address! |
|                 |
|  Cheers, Juhan  |
+-----------------+