Oracle9i RAC Additional SGA Areas

GES/GCS Area -- Global Resource Directory -- The data structures associated with global resources. It is distributed across all instances in a cluster. This area is located in the variable or shared pool section of the SGA. Figure 4.8 shows these additions.

Figure 4.8: Additional Instance Structures for Oracle9i RAC

These processes are listed on the UNIX environment as shown in Figure 4.9.

Figure 4.9 - Example UNIX ps -ef |grep oracle Oracle9i RAC Process List

As you can see from the diagram in Figure 4.8 and the list of processes in Figure 4.9, the Oracle9i RAC environment is a complex set of processes and interrelated memory areas. The memory areas for RAC are taken from the shared pool area. A list of the various components for the SGA can be found by querying the v$sgastat. The combined results from a non-RAC (but with data guard), and with-RAC instance set are shown in Figure 4.10.

Figure 4.10: Comparison between Non-RAC and RAC v$sgastat

The major areas are in the GCS/GES sections shown in the RAC side of the table. These make up the Global Services Dictionary. They are spread amongst the RAC instances and are used to track the global enqueues (locks).

The cache coherency (the latest phase of cache fusion first released in Oracle8i) feature implicit in the Oracle9i RAC configuration is maintained through the Global Services Directory and the various Global Enqueue processes.

Cache coherency is fully implemented in Real Application Clusters. The Oracle9i phase eliminates the previous problems with forced disk writes (refer back to the discussion on pinging). In Oracle9i RAC, instance level fixed locks have been eliminated in favor of intra-instance transportable locks. Cache coherent concurrency control enables excellent performance and scalability for almost all applications.

In Oracle9i RAC, the dynamic resource re-mastering feature is implemented. Dynamic resource re-mastering is the ability of the GCS to move the ownership of a resource between instances of Real Application Clusters during runtime without affecting availability through the high speed interconnect.

This dynamic resource re-mastering is also known as resource affinity. Resource affinity is the use of dynamic resource re-mastering to move the location of the resource masters for a database file to the instance where operations are most frequently occurring. This reduces the traffic on the cluster interconnect and improves performance and scalability.