6.2 Upgrading Nodes in the Cluster (Rolling Cluster Upgrade)

Performing a rolling cluster upgrade lets you keep your cluster up and running and lets your users continue to access cluster resources while the upgrade is being performed.

During a rolling cluster upgrade, one OES server is upgraded at a time to the next SP release. You can also add new nodes running the next SP release. You should complete the upgrade as soon as possible. Don’t leave the cluster in a mixed-mode state for an extended period.

To perform a rolling cluster upgrade:

  1. Make a note of the OES components that are installed on the existing cluster nodes.

    You will probably want to install the same OES components on any new nodes that you add to the cluster.

  2. If you are adding new nodes, prepare the nodes for the cluster, but do not allow them to join the cluster.

    1. Install SUSE Linux Enterprise Server (SLES) 11 SP4 and the OES Services from the OES 11 SP3 Add-On DVD, including Novell Cluster Services, but do not configure the node for the cluster at this time.

    2. Verify that the server is functioning properly.

    3. In YaST, go to the Novell Cluster Services Configuration page as described in Section 5.5.4, Accessing the Novell Cluster Services Configuration Page in YaST.

    4. Configure the node for the existing OES cluster as described in Section 5.5.6, Adding a Node to an Existing Cluster. Complete the following fields:

      Parameter

      Action

      LDAP administrator user password

      Specify the password of the user name shown, then click OK.

      Typically, this is the LDAP administrator identity that was used when the OES Services were installed, or the identity configured after the install in Section 5.5.3, Using Different LDAP Credentials for the Cluster Configuration.

      New or existing cluster

      Select Existing Cluster.

      Directory Server Address

      Select the check box next to the IP address of the LDAP server you want to use as the default for this node. The local LDAP server is selected by default.

      You can add, remove, or change the order of available LDAP servers for the node after the setup is complete. See Section 8.13.1, Changing the Administrator Credentials or LDAP Server IP Addresses for a Cluster.

      Cluster FDN

      Click Browse, navigate the tree to select the existing Cluster object for the OES 11 cluster, click OK, then click Next.

      NCS Proxy User Name

      Specify one of the following users as the NCS Proxy user for this server, then click Next:

      • OES Common Proxy User (default if it was configured for the server)

      • LDAP Admin User

      • Another Administrator User

      IP address of this node

      From the drop-down list, select the IP address that Novell Cluster Services will use for this node.

      Start Cluster Services now

      Deselect the check box.

      Do not start cluster services on this node or join the cluster until Step 4, after all of the Linux POSIX cluster resources have been taken offline.

    5. Click Finish, review the summary, then click Next to complete the configuration.

    6. Click Finish to close the OES Configuration dialog box.

  3. For each of the new nodes in turn, start Novell Cluster services manually by using one of the following methods, and allow the node to join the existing cluster.

    • Reboot the cluster server.

    • At a command prompt, go to the /etc/init.d directory, then enter the following as the root user: 

      ./novell-ncs start

    • At a command prompt, enter the following as the root user:

      rcnovell-ncs start

  4. For each new node, ensure that you modify the Preferred Nodes list accordingly for each of the cluster resources.

  5. For each of the existing cluster nodes in turn:

    1. Log in to the node as the root user, then open a terminal console.

    2. Use the cluster leave command to remove the node from the cluster.

      Any cluster resources that were running on the server should fail over to another server in its preferred nodes list.

      You can also manually cluster migrate the resources to another server in the cluster prior to bringing down the server.

    3. Stop Novell Cluster Services by entering

      rcnovell-ncs stop

    4. If the autostart of Novell Cluster Services is enabled, you must disable autostart in order to prevent it from automatically loading Novell Cluster Services during the upgrade. Enter

      chkconfig -s novell-ncs off

    5. Apply the next SP release as an upgrade, and fully patch the system.

    6. (Optional) Enable the autostart of Novell Cluster Services for the node, enter

      chkconfig -s novell-ncs on

    7. Start Novell Cluster Services by entering

      rcnovell-ncs start

    8. Use the cluster join command to rejoin the node to the cluster.

    9. Modify the Preferred Nodes list accordingly for each of the cluster resources.

    10. Repeat these steps for each existing node in the cluster.

  6. After all of the existing nodes have been upgraded and any new nodes added, the conversion is complete.