prime dock
We have a NFS share for storing Oracle backups and with tiered "auto" policy and 7 days tiering-minimum-cooling-days. We have found that when we restore more than 7 days older data, it will take extremly longer time due to the extra time taken for retrieving older data from StorageGrid.
Based on my understanding, sequential read would not cause data retrieval, and Oracle RMAN restore is sequential read(correct?) , why did it actually retrieve data to local storage from SG?
We know it should take longer because the time on retrieving from the object storage, usually about 10-20ms extra in StorageGrid case. But, even consider that, it should not take THAT long. In the end, we have to kill the restore.
mighty rover