Article ID: 117032, created on Aug 30, 2013, last review on May 10, 2014

  • Applies to:
  • Virtuozzo

Symptoms

  1. A resource (container or virtual machine) is marked as broken in shaman stat :

    ~# shaman stat | grep Broken
    50                    running     Broken      Unknown           0
    
  2. The resource is now running fine on one of the nodes in the cluster.

Cause

Shaman resource experienced relocation failure in the past and had been marked as broken.

NOTE: If there are issues with the resource still and it is not functioning correctly, resolve all the issues first.

Resolution

For containers:

  1. Suspend the container:

    ~# vzctl suspend 50
    
  2. Unregister it:

    ~# vzctl unregister 50
    
  3. Delete the related resource file from the shaman directory with broken resources:

    ~# rm -f /pstorage/cluster_name/.shaman/broken/ct-50
    
  4. Register the container back:

    ~# vzctl register /vz/private/50 50
    
  5. Resume it:

    ~# vzctl resume 50
    

For virtual machines:

  1. Suspend the affected VM:

    ~# prlctl suspend MyVM
    
  2. Unregister the VM:

    ~# prlctl unregister MyVM
    
  3. Delete the related resource file from shaman directory with broken resources:

    ~# rm -f /pstorage/cluster_name/.shaman/broken/vm-MyVM
    
  4. Register the virtual machine back:

    ~# prlctl register /pstorage/cluster_name/vmprivate/MyVM.pvm/config.pvs --preserve-uuid
    
  5. Resume it:

    ~# prlctl resume MyVM
    

Search Words

shaman stat

shaman

broken

relocation failed

0dd5b9380c7d4884d77587f3eb0fa8ef 2897d76d56d2010f4e3a28f864d69223

Email subscription for changes to this article
Save as PDF