This topic does not apply to Oracle Cloud Infrastructure or to Oracle Cloud at Customer.
Oracle RAC enhances Oracle Database capabilities so that you can concurrently use multiple database instances on different compute nodes. This allows you to scale workload across multiple database instances in order to efficiently store, update, and retrieve data.
Oracle RAC provides the software that manages multiple servers and database instances as a single set of servers, called a cluster. The data files that comprise the database reside on shared storage that is accessible from all servers that are part of the cluster. Each server in the cluster runs the Oracle RAC software.
Unlike a single-instance Oracle database, which has a one-to-one relationship between data files and the database instance, Oracle RAC databases have a one-to-many relationship between data files and database instances. This means that in an Oracle RAC database multiple database instances access a single set of database files concurrently, allowing you to access the data from any database instance in the database cluster.
This allows you to use horizontal scalability beyond the scope of one compute node, in case this compute node is insufficient to run the desired workload. It also increases availability of the database and the data in case a database instance or compute node fails. The remaining database instance can be used to continue operations while the failed database instance or compute node is being restarted. Having more than one database instance also allows you to perform rolling patch upgrades.
To learn about using Oracle Real Application Clusters and Oracle Data Guard together, see Using Oracle Real Application Clusters (RAC) and Oracle Data Guard Together in Database Classic Cloud Service.
To create an Oracle RAC database in Database Classic Cloud Service, make the following choices in the Create Instance wizard:
For Software Edition, choose Enterprise Edition - Extreme Performance.
For Database Type, choose Database Clustering with RAC.
For Compute Shape (on the Service Details page), choose a shape with two or more OCPUs.
When you make these choices, Database Classic Cloud Service creates a two-node Oracle RAC database, hosting the database on two independent compute nodes that share data, fast recovery area, and redo log storage. It creates these compute nodes using computing, storage and networking resources provided by Oracle Cloud Infrastructure Compute Classic.
Managing a Deployment that Uses Oracle RAC Compared to a Deployment that Doesn’t
Because a Database Classic Cloud Service deployment that uses Oracle RAC comprises two compute nodes that each host a RAC database instance, you manage the deployment in slightly different ways:
Cloud tooling: you use
dbaascli. For more information, see The raccli Utility.
You can stop and start the database instances and even the compute nodes independently of each other. Thus, the database can remain available even when you need to perform maintenance that requires you to stop a database instance or compute node.
More About the Oracle RAC Configuration on Database Classic Cloud Service
Cluster size: currently, the Oracle RAC database on a Database Classic Cloud Service deployment is limited to a two-node cluster.
Cloud tooling: Oracle Cloud tooling is provided for the common administrative tasks of scaling, backing up and recovering, and patching. For more information, see Scaling a Database Deployment and The raccli Utility.
Network access: on the compute nodes, access to all ports except port 22 is disabled. Port 22 is open for passwordless, key-based SSH access by the
opcuser. To enable access to other ports, see Enabling Access to a Compute Node Port.
Networking for client access: to make client connections to the Oracle RAC database, you include particular options in the connection’s entry in the client’s
tnsnames.orafile. For more information, see Connecting Remotely to the Database by Using Oracle Net Services.
Database file storage: storage for database data files, the fast recovery area, and the redo logs is created and managed using Oracle Automatic Storage Management (ASM) and Oracle Automatic Storage Management Cluster File System (ACFS) instead of Linux LVM.
Included software: Oracle Grid Infrastructure, Oracle ASM and Oracle ACFS are included; Oracle Application Express, Oracle REST Data Services, and Oracle SQL Developer Web are not currently included.