For example, do one of the following tasks:

• Roll-forward recovery (or replay all log files)

  After you restore the files and the service starts up, Exchange commits the transactions in the logs you restored. If contiguous logs exist on the server beyond the log with the highest number you restored, those transactions also are committed. If there is any gap in the numeric sequence of log names, no further transactions are committed beyond the gap.

  This scenario is useful when the transaction logs are intact but you require the database to be restored. When you keep existing transaction logs, Exchange Server can recover to the point of the failure. Otherwise, you must recover to the time of the last full backup or the last incremental backup.

• Point-in-time recovery (or replay only restored log files)

  Use this option if you only want to restore up to the point of the last backup. Any transaction logs that are created after the last backup are not involved in the recovery of the database(s). For snapshot restores, NetBackup deletes the current log files.