I have an interesting problem. I am running full backup cycles of the database twice a week and we have re-directed the backuppieces to a DataDomain device. THe problem is the performance has tanked. Normal backups to a NAS take approximately 5 hours. The *new and improved* DataDomain option takes just under 16 hours. I have modified the rman script to conform to DataDomain's Best Practice docs (i.e...filesperset=1, uncompressed, yada...yada...) but I still am unable to achieve acceptable performance. The database configuration follows Oracle's MAA whitepaper on High Availability. I use a local standby to run the full backup to offload the CPU and memory burden from the production environment. I run semi-annual tests on the backups to verify they are indeed good and I've never hade any issues. We run nightly incrementals on the primary using block change tracking and image copies of the datafiles. Again, never have had any issues. Can anyone offer any suggestions?