PRIMECLUSTER Installation and Administration Guide 4.5

I.5.1 Backup/Restore of Virtual Machine by Snapshot Function

When backing up or restoring the virtual machine by using the snapshot function of RHOSP, take the following steps.

I.5.1.1 Backing up Virtual Machine

1. If GDS is used, set up according to the following manual.

   See

   Refer to "Settings Before Backing Up" of "Backing Up and Restoring System Disk" in "PRIMECLUSTER Global Disk Services Configuration and Administration Guide."

2. Create the snapshot of the virtual machine.

   See

   For how to create snapshots, refer to the RHOSP manual of Red Hat, Inc.

3. If GDS is used, set up according to the following manual.

   See

   Refer to "Settings After Backing Up" of "Backing Up and Restoring System Disk" in "PRIMECLUSTER Global Disk Services Configuration and Administration Guide."

I.5.1.2 Restoring Virtual Machine

In either of the following cases, take the following steps for restoring:

• Stop all the nodes before restoring.

• After restoring one node during business operation, fail over the active node and restore the other node.

  
  

[How to restoring]

1. If GDS is used, set up according to the following manual.

   See

   Refer to "Settings Before Restoring" of "Backing Up and Restoring System Disk" in "PRIMECLUSTER Global Disk Services Configuration and Administration Guide."

2. Restore the virtual machine from the snapshot.

   1. For the virtual machine to be restored, check the virtual machine type and the additional volume ID.

   2. Delete the virtual machine to be restored.

   3. Perform "1. Creating Port for Public LAN (also used as administrative LAN)" and "2. Creating Port for Cluster Interconnect" to create the port.

   4. Restore the virtual machine from the snapshot. At the same time when restoring, OS is started.

      Set up the virtual machine to be restored as follows.

      Item	Setting value
      Virtual machine name	Any virtual machine name *Do not specify a duplicate virtual machine name in the same project.
      Virtual machine type	Flavor ID of the virtual machine type checked in Step 1
      Keypair name	Key pair necessary for ssh login to virtual machine
      Port ID	Port ID (eth0) created in "1. Creating Port for Public LAN (also used as administrative LAN)" Port ID (eth1) created in "2. Creating Port for Cluster Interconnect"
      Security group	Not specified (already specified in the port)
      Auto-failover	Disabled
      Server group ID	Server group ID created in "I.2.2.3 Creating Server Group"
      Minimum number of servers	1
      Maximum number of servers	1
      Snapshot ID	ID of snapshot
      Additional volume ID	ID of the additional volume checked in Step 1
      Size of additional volume	Size of the additional volume checked in Step 1
      Device path of additional volume	Device path of the additional volume checked in Step 1

      Note

      Make sure to use this procedure to set up the additional volume registered in GDS.

      If the additional volume is not set up during this procedure, do not attach the additional volume to the restored virtual machine but restore the virtual machine again according to this procedure. If the additional volume is attached to the restored virtual machine, the remaining steps fail.

3. If GDS is used, set up according to the following manual.

   See

   Refer to "Settings After Restoring" of "Backing Up and Restoring System Disk" in "PRIMECLUSTER Global Disk Services Configuration and Administration Guide."

4. If the virtual machine name has been changed in Step 2, take the following steps and changed the settings of shutdown facility.

   1. Execute the following command on all the nodes to stop the shutdown facility.

      # sdtool -e

   2. Describe the changed virtual machine name to the configuration file of the shutdown agent.

      See

      For the descriptions of configuration file, refer to "2. Setting up Shutdown Facility."

   3. Execute the following command on all the nodes to start the shutdown facility.

      # sdtool -b

   4. Execute the following command on all the nodes and make sure that the shutdown facility operates normally.

      # sdtool -s

   Note

   There is a possibility that the settings of the agent or network are not correct when any of the following statuses are displayed though changing the settings of the shutdown facility is completed.

   • InitFailed is displayed in Init State.

   • Unknown or TestFailed is displayed in Test State.

   In this case, review the settings of the agent or network.