Host CPU does not provide required features | Nutanix Community
Skip to main content

I have several Citrix clusters with NX-8170-G8 and NX-8170-G9 nodes.  I want to perform updates but the VMs will not migrate. VMs need to be power cycled in order to get the VMs revert to the lower CPU in the cluster.

  1. I was looking for script that i can run in the intelligent ops playbook that will Identify the VMs that needs to be power cycled, and power cycle them. 
  2. Can assist with script to get the state of the VMs that needs to be power cycled? 
  3. Power cycle the VMs which don't have APC configured. 
    Power cycle the VMS which don't have APC configured to ensure CPU compatibility.
    Power cycle all the VMS in the cluster to ensure CPU compatibility. 

How do I find VMs that do not have APC configured? This one of the VMs affected that needs to be power cycled.


{
  can_clear_removed_from_host_uuid: True
  config {
    agent_vm: False
    allow_live_migrate: True
    annotation: "Updated 4/14/23 to match Xen Snapshot 35"
    apc_config {
      apc_enabled: False
    }
    bios_uuid: "71ca6758-0ee3-49d9-8bec-3f3f046bbdfa"
    boot {
      boot_device_order: "kCdrom"
      boot_device_order: "kDisk"
      boot_device_order: "kNetwork"
      hardware_virtualization: False
      secure_boot: False
      uefi_boot: False
    }
    cpu_passthrough: False
    disable_branding: False
    disk_list {
      addr {
        bus: "scsi"
        index: 0
      }
      cdrom: False
      container_id: 2262
      container_uuid: "b6043504-0b7f-46d0-b2c3-23d30282020a"
      device_uuid: "026bab2e-dc4e-4d65-8ea5-95aaaa61bdf9"
      naa_id: "naa.6506b8db23064a5308489ba16fe2a649"
      source_vmdisk_uuid: "f03ac63a-6e68-4ed6-971c-3932602db3e4"
      storage_vdisk_uuid: "da730843-d298-42b0-83fe-258317ccd547"
      vmdisk_size: 118111600640
      vmdisk_uuid: "10487252-a0af-4448-9591-d9383744b13f"
    }
    disk_list {
      addr {
        bus: "ide"
        index: 0
      }
      cdrom: True
      device_uuid: "edfe70e5-94d1-4a6e-b3f8-43d6d2c32f30"
      empty: True
      iso_type: "kOther"
    }
    disk_list {
      addr {
        bus: "scsi"
        index: 2
      }
      cdrom: False
      container_id: 2262
      container_uuid: "b6043504-0b7f-46d0-b2c3-23d30282020a"
      device_uuid: "d8e6958c-e526-4fa6-bc22-7f52ded26205"
      naa_id: "naa.6506b8d3dcd56efe9cda657bb513cf0f"
      source_vmdisk_uuid: "c0efb44a-8826-4bc6-a35e-ac1f18a53481"
      storage_vdisk_uuid: "cff75a61-882b-4c4c-8551-8df251665a81"
      vmdisk_size: 16777216
      vmdisk_uuid: "a2448090-12ac-45b2-9afb-3ea151ba0df8"
    }
    flash_mode: False
    generation_uuid: "700bc578-8d27-45d7-ba50-e61e4b2a0f9a"
    gpu_console: False
    hwclock_timezone: "America/Chicago"
    machine_type: "pc"
    memory_mb: 24576
    memory_overcommit: False
    name: "CTX-GEN015"
    ngt_enable_script_exec: False
    ngt_fail_on_script_failure: False
    nic_list {
      connected: True
      mac_addr: "50:6b:8d:a8:04:5a"
      network_name: "aaaaaa"
      network_type: "kNativeNetwork"
      network_uuid: "e597890c-702c-4111-9f31-c08e087141b2"
      queues: 1
      rx_queue_size: 256
      type: "kNormalNic"
      uuid: "78382445-d4a3-4758-9cb6-92b8115d6d21"
      vlan_mode: "kAccess"
    }
    num_cores_per_vcpu: 2
    num_threads_per_core: 1
    num_vcpus: 3
    num_vnuma_nodes: 0
    power_state_mechanism: "kHard"
    vcpu_hard_pin: False
    vga_console: True
    vm_type: "kGuestVM"
    vtpm_config {
      is_enabled: False
    }
  }
  cpu_model {
    name: "Intel Icelake"
    uuid: "2c7c5833-3f05-5960-8c6b-18e880fdfc71"
    vendor: "Intel"
  }
  cpu_model_minor_version_uuid: "106d6914-d715-569f-b663-b4cb095ba424"
  host_name: "10.0.29.214"
  host_uuid: "b0778f6e-4b3e-4d56-9b25-b9960f9bc10a"
  is_ngt_ipless_reserved_sp_ready: True
  is_rf1_vm: False
  logical_timestamp: 593
  protected_memory_mb: 24576
  state: "kOn"
  uuid: "71ca6758-0ee3-49d9-8bec-3f3f046bbdfa"
}
 

Hi Pathfinder,

VMs must to be power off because you’re using CPU passthrough ?

Regards,

How can I see or find where I am using CPU Passthrough? Apologies for the ignorance on this.

 

How can I see or find where I am using CPU Passthrough? Apologies for the ignorance on this.

 

Oh sorry i misunderstood, i seed your post edited.

With the latest Prism Central you can give vm’s a specific feature set. Set this to the lowest in the cluster and you are good to go. 
 

 

The problem is I have to power down VMs to make that change for each VM. On the other hand, If I power cycle the VMs, as KB 6409 suggest as follows then my problem is solved. These are Citrix VMs. 

The VMs must be power cycled to get the correct CPU topology and feature set. A full power off and power on of the VM is required to resolve this issue. A reboot is not sufficient. 
Shut down the affected VMs on the list.
Start the VMs from Prism UI.

Yes, when you have different types of chipset nodes in a cluster and the vm's where already running you need to reboot them. (power off and on). 

But as this are Citrix machines I assume you are using MCS so the reboot (shutdown/power on) will be done by Citrix automatically for the image update or reboot or something else. 

 

Reply


Terms & Conditions