Procedure for "Snapshot" or "Revert Snapshot" of a Virtual Machine while VCS for VMware is online

Article:TECH146913  |  Created: 2010-12-23  |  Updated: 2010-12-24  |  Article URL http://www.symantec.com/docs/TECH146913
Article Type
Technical Solution

Product(s)

Environment

Issue



Executing "reverting snapshot of VMware Virtual Machine" while VCS is online GuestOSApp status goes unknown. The problem will only occur on VCS for VMware. 

 


Error



VCS WARNING V-16-10061-16757 GuestOSApp:<GenericService Resource Name>:
monitor:VCS virtual machine tools is not reporting the resource's state inside of the virtual machine.
 


Cause



On reverting the Virtual Machine snapshot, VCS unable get resource status as it is cleared.
Resulting ESX Server and Virtual Machine failed to communicate anymore and hence the resource goes into "unknown" state.

 


Solution



 The snapshot and revert operation of a Virtual Machine is outside VCS operation but it directly affects VCS process inside the GuestOS. Therefore, it is required to stop vcsagmd before proceeding with snapshot or revert a machine.

A)If you preparing for snapshot or revert a machine.
(Also See "Veritas Cluster Server Implementation Guide" -> "Performing maintenance on applications inside the virtual machine")

  1. For Linux, open a terminal window, find the home/install directory of the VCS Virtual Machine Tools, and make sure it is in the PATH.
    # export PATH=$PATH:/opt/VRTSvcs/bin

  2. Find the VCS resource name and the service group name that is associated with this virtual machine.
    To find the resource name on Windows, type the following command:
    C:\> type "%VCS_HOME%\.vcsvmresname"

    To find the resource name on Linux, type the following command:
    # cat /etc/VRTSvcs/.vcsvmresname
    ■ To find the service group that the resource resides in, type the following command at the prompt:
    hares -value vcs_vm_res_name Group

    Where vcs_vm_res_name is the name of the resource that you found.
    Record the output of the hares command, you need to use this service group name in the following step.

  3. Freeze the service group. At the prompt, type:
    # hagrp -freeze service_group_name

    Where service_group_name is the name of the service group that the 
    resource resides in.

  4. Stop the VCS agent management daemon. At the prompt, type:
    On Windows:
    C:\> net stop vcsagmd
    On Linux:
    # /etc/init.d/vcsagmd stop

  5. Perform the Virtual Machine Snapshot operation (Taking/Reverting snapshot)
     
  6. Start the VCS agent management daemon. At the prompt, type:
    # /etc/init.d/vcsagmd start

  7. Unfreeze the service group. At the prompt, type:
    # hagrp -unfreeze service_group_name

    Where service_group_ group_name is the name of the service group that the 
    resource resides in.


B) If you have already observed the problem, restart the VCS agent management daemon(vcsagmd) within the GuestOS.
    - On Windows:
       C:\> net stop vcsagmd
       C:\> net start vcsagmd
    - On Linux:
       # /etc/init.d/vcsagmd stop
       # /etc/init.d/vcsagmd start
 




Article URL http://www.symantec.com/docs/TECH146913


Terms of use for this information are found in Legal Notices