Applying Patches to Oracle Clusterware Using DB Control

To patch the Oracle Oracle Grid Infrastructure home, all applications that depend on Oracle Clusterware will need to be stopped for the node being patched. In the case of a non-shared Oracle Oracle Grid Infrastructure home, the patch is applied in a rolling method, so only a small part of the cluster is down at any given time. If you have a shared Oracle Oracle Grid Infrastructure home, then you can use the out-of-place patching method to minimize the downtime required for patching.

If you want to patch the software for ASM, because it uses the same directory as Oracle Clusterware you must also shut down Oracle Clusterware prior to patching.

This section describes the procedure for patching a non-shared Oracle Oracle Grid Infrastructure home. The patches are applied in-place, that is, to the existing software directories.

To patch Oracle Oracle Grid Infrastructure home using DB Control, perform the following steps:

1. In DB Control, click the link Deployments.

   for the Deployment Procedures page, then click Patch Oracle Clusterware - Rolling.

   A list of patches currently available in the Software library is displayed, or you can click Add Patches to locate and download the patch from My Oracle Support (formerly OracleMetaLink).

2. DB Control automatically selects the current Oracle Clusterware installation to be patched. Select all the nodes that need to be patched.

3. Enter the operating system credentials for the Oracle Oracle Grid Infrastructure home being patched.

4. (Optional) Specify a schedule for when you want the patching operation to occur. If you want the patch to be applied immediately, skip this step.

5. Review the information about the selections you have made. If you need to make changes to the information, use the Back button.

6. DB Control starts the patching operation. It performs the following tasks:

   • Downloads patches or patchsets from My Oracle Support and stages them on the local file system

   • For one of the nodes in the cluster, all databases and cluster database instances are stopped, the ASM instance is stopped, and any other Oracle Clusterware dependent resources are stopped. The unroot.sh script is run on that node, if required by the patch. The patches are then applied to the Oracle Oracle Grid Infrastructure home. If necessary, the root.sh script is run for the Oracle Oracle Grid Infrastructure home, and then Oracle Clusterware is restarted, followed by all the dependent resources that were shut down at the start of the patching operation.

   • The previous step is repeated for each node in the cluster