Oracle Real Application Clusters

12.1 Introduction

In the previous chapter we discussed the various aspects considered during the migration process. There are three possible migration scenarios:

1. Migrating from single instance to RAC.

2. Migrating from OPS to RAC.

3. Migrating from single instance to OPS.

Migrating a database from a single instance configuration to a RAC configuration is more complex compared to the migration from an OPS environment to RAC. However, the earlier migration (scenario 3 above, which was not discussed in the previous chapter) of a system from a stand- alone configuration to OPS was comparatively much more complex and involved a great amount of planning and architectural reconfiguration. This is primarily because of the complexity that was present with users sharing information from multiple instances, and the high pinging activity that is not present under Oracle 9i.

In this chapter we will discuss backup and recovery in a RAC environment. Backups are taken to protect against the future, that is, to protect against failure situations that could potentially happen. One may argue that the chances of these failures happening are 50–50; however, when it happens without these backups it would be impossible to get back the lost data. Once it is certain that backup is required, the next question is, how often are they to be backed up? This depends on the backup strategy defined for the organization.

When we mention recovery, it also covers failures of an instance or system, because Oracle has to recover from these failures. Availability of an Oracle database when an instance or node fails, has been discussed in detail in Chapter 10 (Availability and Scalability). This chapter, apart from the backup functionality, will discuss in detail the various failure scenarios and the recovery process during those failures.