Based on reading some documentation on Metalink, the answers to your questions are
1. RMAN doesn't require a large pool allocation, unless you have i/o slave processes enabled by question 2. Of course you may still need a large pool allocation for other reasons.
2. The i/o slave processes that are relevant to RMAN are backup_disk_io_slaves and backup_tape_io_slaves. One reason for using these parameters is if you are running on an O/S that either doesn't support asynchronous i/o, or implements it inefficiently. The goal would be to improve throughput of your backup jobs.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.