4.1. Understanding the Restore and Recovery Process

There are two main steps required to perform a database recovery: restoring and recovering. These are the same whether you perform a recovery using user-managed recovery procedures, Recovery Manager (RMAN), or Enterprise Manger (EM).

The restore process consists of copying the required physical files that are needed to begin the recovery of the database. If user-managed recovery is performed, the restore process will be a manual process that is based on operating systems commands or third-party tape management GUIs that help you retrieve the files from tape or disk and place the files in the appropriate locations. If using RMAN or EM to recover, this process will be performed by issuing the RESTORE command. The RESTORE command retrieves the required files from the backup media tape or disk and places the files in the appropriate location.

Once the files have been restored, the recovery process can begin. The recovery process is the process of taking a datafile and applying archived redo logs, which apply the changes that occurred in the database to the datafile.

Figure 4.1 shows the complete restore and recovery process. In this figure, a database backup was performed on August 1. Archive logs were generated from 101 through 322, at which point a database failure occurred on August 4. So the backup from August 1 was restored and archive logs 101 through 322 were applied to the backup to roll the backup forward to the point of the database failure.

Figure 4.1. The restore and recovery process