Create and Deploy on KVM

For complete KVM documentation, refer to http://www.linux-kvm.org/page/Documents.

1. Install the Virtualization Host group and virt-install.

   # yum groupinstall "Virtualization Host"
   # yum install virt-install

2. Extract the image.

   # tar xvf nnSCZ739.64-img-vm_kvm.tar
   nnSCZ739.64-img-vm_kvm.ovf
   nnSCZ739.64-img-vm_kvm.qcow2
   legal.txt

3. Use virt-manager to create the management and media network interfaces.
   • Create a virtual bridged network for management interfaces.
   • Create virtual networks for media interfaces.
4. Provision a new virtual machine.

   # virt-install \
       --name SBC739 \
       --description "nnSCZ739 KVM" \
       --os-type=Linux \
       --os-variant=rhel7 \
       --ram=8192 \
       --vcpus=4 \
       --disk path=/opt/nnSCZ739.64-img-vm_kvm.qcow2,bus=virtio,size=10,format=qcow2 \
       --network bridge=br-Mgmt \
       --network bridge=br-Mgmt \
       --network bridge=br-Mgmt \
       --network bridge=br-Mgmt \
       --network network=media1 \
       --network network=media2 \
       --import \
       --cpu host

   Note:

   Use interface-mapping to pin the four br-Mgmt network interfaces to wancom0, wancom1, wancom2, and spare.

   --name

   Identify a unique name for the virtual machine on this hypervisor.

   --description

   Describe this virtual machine.

   --os-type

   Specify the operating system type.

   --os-variant

   Optimize the configuration for a specific operating system.

   --ram

   Allocate a specific amount of RAM to this virtual machine.

   --vcpus

   Allocate a specific number of virtual CPUs to this virtual machine.

   --disk

   Specify the path to the disk image.

   --network

   Connect the virtual machine to a host network.

   --import

   Skip the operating system installation process and build a guest around the disk image specified with --disk.

   --cpu

   Configure the CPU model and CPU features exposed to the virtual machine.

   See man virt-install for more information.

   Note:

   The --cpuset and --numatune options may be used to establish CPU affinity and socket pinning.

To maintain performance when using PV mode, you must pre-allocate cores to the Open vSwitch (OVS) for each forwarding core in the VM. For example, if the VM requires four forwarding cores, then four CPU cores should be allocated to the OVS.