Installing SF Oracle RAC on the secondary site
Important requirements for global clustering:
-
Cluster names on the primary and secondary sites must be unique.
-
Node and resource names must be unique within a cluster but not across clusters.
You can set up a multi-node or single-node cluster on the secondary site. The only difference between setting up a single node-cluster and a multi-node cluster is the fencing configuration.
See Installing SF Oracle RAC on a multi-node cluster
See Installing SF Oracle RAC on a single-node cluster
Installing SF Oracle RAC on a multi-node cluster
If you install SF Oracle RAC on a multi-node cluster on the secondary site, you can enable I/O fencing feature of SF Oracle RAC.
To install SF Oracle RAC on a multi-node cluster
-
Prepare to install SF Oracle RAC by planning your installation, reviewing the requirements, and performing pre-installation tasks.
See Planning SF Oracle RAC installation and configuration
See Preparing to install and configure SF Oracle RAC
-
Install SF Oracle RAC.
-
Set up storage on the secondary site and configure I/O fencing.
-
Shut down and restart all nodes on the secondary site.
-
After starting the cluster nodes, check that the cvm group is online:
# hagrp -state cvm
Installing SF Oracle RAC on a single-node cluster
For a single-node cluster, after installing and configuring SF Oracle RAC, do not enable I/O fencing. You must not create a coordinator disk group and edit the configuration files to enable fencing. Fencing will run in disabled mode.
If you add more nodes later to the existing single-node cluster on the secondary site, then you must enable I/O fencing.
See Adding a node to an Oracle9i cluster
See Adding a node to an Oracle 10g cluster
To install SF Oracle RAC on a one-node cluster
-
Prepare to install SF Oracle RAC by planning your installation, reviewing the requirements, and performing pre-installation tasks.
See Planning SF Oracle RAC installation and configuration
See Preparing to install and configure SF Oracle RAC
-
Install SF Oracle RAC.
-
Set up storage on the secondary site.
-
Shut down and restart the node on the secondary site.
-
After starting the cluster node, check the cvm group is online:
# hagrp -state cvm