We have a new 19c database (18.104.22.168) with approximately 25TB of data. There are two bigfile tablespaces in this database and 99% of the data are in these two tablespaces. I used the following script to make a full backup:
The problem was that the backup would take approximately 60 hours to complete. That's a lot, and when I checked the rman progress I noticed that sixteen backup channels were released in the first thirty minutes, leaving only two slow channels!
Therefore, I modified the script as below and decided to backup those two tablespaces separately:
After running the backup I checked the rman progress again and saw that the parallel channels were working fine and the backup finished within four hours.
My question is that I haven't had this problem with bigfile tablespaces in earlier databases (like 12c or even 19.12). Thank you for your opinion on this matter.