This process is based on several interventions that can run on different days in order to reduce the downtime of the Environment. The main determinant of this action is that there is no storage available to house two databases simultaneously and need to reconfigure all disks and diskgroups of the ASM instance.
Operating System Upgrade to RHEL 5.7
This action requires the disconnection of all users and must run the Systems Area following the procedure carried out in the Development RAC Cluster with version 126.96.36.199.
Once the operating system upgrade and re-configuration of disk arrays (RDAC) systems intervention is terminated.
Then we proceed to re-link all the software binaries Clusterware and Oracle© Database and restart services.
Oracle Software Update to 188.8.131.52
This action may be initiated immediately after the previous intervention and is divided into the following phases:
Re-Installation Software Node 1
Without interrupting the database service access is necessary we proceed to stop the first of the two nodes of the cluster and de-install all Oracle© software (Clusterware and Database).
Re-Install the Clusterware and Database Software to the new version (184.108.40.206) and application patches. Configuration and parameterization as implemented in the new development specifications.
Creation and configuration of basic database to prepare the transfer of data.
We proceed to the daily export the entire contents of the database lasting approximately 2 hours (in Node 2). For this next action is necessary to stop the database service, so that would take place late in the afternoon schedule.
Stops the second node, and delete the database and it will be necessary to reconfigure the ASM storage. Then dumpfiles are imported to the new database from the export.
We proceeded to inspect the results and the start Node 1 for providing the database service again.
Cluster Phase Completion
With first Node in service, all Oracle@ Clusterware and Database Software is uninstalled from second node.
Finally run the the node clonation from Node 1 to Node 2. This procedure can be performed late in the afternoon to avoid interruption of service database for an incident.