Hi,
Why does the first file part in a backupset take a long time to restore proportionate to the size of the tablespace whilst the subsequent file parts transfer in a matter of minutes?
We have a 3.7TB tablespace backed up as a backupset and split into parts of roughly 100Gb. As part of a RMAN Duplicate, when the first part is allocated to a channel everything seems to stall (little Network, SAN, or disk usage) until RMAN has allocated all the space on ASM. Once the first part finishes all the subsequent backupset parts transfer in roughly 15 minutes.
The has caused problems as our Backup solution - HP StoreOnce has a timeout of 4 hours and the first part of this backup set takes >4 hours to restore. Increasing the timeout to 6 hours was a temporary work around but I’d like to understand whats going on for this first file part and if there is anyway to reduce the time its taking.
Oracle 11.2.0.4, RAC and ASM
OS: AIX 7.1
Backup media – HP StoreOnce, with HP StoreOnce Catalyst plugin for RMAN, HP
Thanks
S