Chapter 11. Implementing Database Recovery

ORACLE DATABASE 10 G: ADMINISTRATION I EXAM OBJECTIVES COVERED IN THIS CHAPTER:

• Backup and Recovery Concepts

  • Describe the types of failure that may occur in an Oracle Database.

  • Describe ways to tune instance recovery.

  • Identify the importance of checkpoints, redo log files, and archived log files.

• Database Recovery

  • Recover from loss of a Control file.

  • Recover from loss of a Redo log file.

  • Recover from loss of a system-critical data file.

  • Recover from loss of a non system-critical data file.

Oracle Database 10g (Oracle 10g) makes it easy for you to recover from a number of database failures. In Chapter 10, "Implementing Database Backups," we emphasized the importance of checkpoints, redo log files, and archived log files to maintain a high level of availability and recoverability. We also showed you several ways to back up your database. In this chapter, we'll show you how to use those backups effectively when some kind of failure inevitably occurs.

First, we'll review the kinds of failures that can occur in the database; they can occur because of mistakes by users or DBAs or because of hardware or software failures that are out of your direct control. Each of these failures can require little or no action whatsoever, as in the case of an instance failure, but at the other end of the spectrum, a crash of the disk containing the SYSTEM tablespace requires a recovery effort.

We'll then emphasize the importance of checkpoints, online redo log files, and archived redo log files. To balance performance with recoverability, we'll also show you how to tune instance recovery to minimize the amount of time Oracle will require to recover from an instance failure while still providing a reasonable response time for ongoing transactions. In a nutshell, your job is to increase the mean time between failures (MTBF) by providing redundant components where possible and leveraging other Oracle high-availability features such as Real Application Clusters (RAC) and Streams (an advanced replication technology). Going hand in hand with MTBF is decreasing the mean time to recovery (MTTR) to ensure compliance with service-level agreements that are in place. Last, but certainly not least, these efforts should help to minimize data loss in such a way that committed transactions are never lost.

Near the end of this chapter, we'll review the steps required to recover from the loss of both system critical and non–system-critical datafiles for databases that are operating in both ARCHIVELOG and NOARCHIVELOG mode. We'll also show you how to recover from the loss of a control file or a redo log file. As with most DBA operations in the database, Oracle's GUI administration tool, the Enterprise Manager (EM) Database Control, makes many of these administration tasks easier and less error prone.