Storage Foundation for Oracle® RAC 7.3.1 Administrator's Guide - Linux

To migrate from disk-based fencing to server-based fencing using the installer

1. Make sure system-to-system communication is functioning properly.
2. Make sure that the SF Oracle RAC cluster is online and uses disk-based fencing.

   # vxfenadm -d

   For example, if SF Oracle RAC cluster uses disk-based fencing:

   I/O Fencing Cluster Information:
   ================================
   Fencing Protocol Version: 201
   Fencing Mode: SCSI3
   Fencing SCSI3 Disk Policy: dmp
   Cluster Members:
   		* 0 (sys1)
   		1 (sys2)
   RFSM State Information:
   		node 0 in state 8 (running)
   		node 1 in state 8 (running)

3. On any node in the cluster, start the installer with the -fencing option.

   # /opt/VRTS/install/installer -fencing

   Note the location of log files which you can access in the event of any problem with the configuration process.

4. Confirm that you want to proceed with the I/O fencing configuration.

   The installer verifies whether I/O fencing is configured in enabled mode.

5. Confirm that you want to reconfigure I/O fencing.
6. Review the I/O fencing configuration options that the program presents. Type 4 to migrate to server-based I/O fencing.

   Select the fencing mechanism to be configured in this
   Application Cluster [1-4,q] 4

7. From the list of coordination points that the installer presents, select the coordination points that you want to replace.

   For example:

   Select the coordination points you would like to remove
   from the currently configured coordination points:
      1) emc_clariion0_62
      2) emc_clariion0_65
      3) emc_clariion0_66
      4) All
      5) None
      b) Back to previous menu
   Enter the options separated by spaces: [1-5,b,q,?] (5)? 1 2

   If you want to migrate to server-based fencing with no coordinator disks, type 4 to remove all the coordinator disks.

8. Enter the total number of new coordination points.

   If you want to migrate to server-based fencing configuration with a mix of coordination points, the number you enter at this prompt must be a total of both the new CP servers and the new coordinator disks.

9. Enter the total number of new coordinator disks.

   If you want to migrate to server-based fencing with no coordinator disks, type 0 at this prompt.

10. Enter the total number of virtual IP addresses or host names of the virtual IP address for each of the CP servers.
11. Enter the virtual IP addresses or host names of the virtual IP address for each of the CP servers.
12. Verify and confirm the coordination points information for the fencing reconfiguration.

13. Review the output as the installer performs the following tasks:

    • Removes the coordinator disks from the coordinator disk group.

    • Updates the application cluster details on each of the new CP servers.

    • Prepares the vxfenmode.test file on all nodes.

    • Runs the vxfenswap script.

      Note the location of the vxfenswap.log file which you can access in the event of any problem with the configuration process.

    • Completes the I/O fencing migration.

14. If you want to send this installation information to Veritas, answer y at the prompt.

    Would you like to send the information about this installation
    to Veritas to help improve installation in the future? [y,n,q,?] (y) y

15. After the migration is complete, verify the change in the fencing mode.

    # vxfenadm -d

    For example, after the migration from disk-based fencing to server-based fencing in the SF Oracle RAC cluster:

    I/O Fencing Cluster Information:
    ================================
    Fencing Protocol Version: 201
    Fencing Mode: Customized
    Fencing Mechanism: cps
    Cluster Members:
    		* 0 (sys1)
    		1 (sys2)
    RFSM State Information:
    		node 0 in state 8 (running)
    		node 1 in state 8 (running)

16. Verify the current coordination points that the vxfen driver uses.

    # vxfenconfig -l