KVM-Specific Configuration

Use these settings to ensure optimal performance and compatibility when deploying vSZ-D on a KVM hypervisor.
Note: For all future releases, refer to RUCKUS SmartZone (LT-GA) Tunnel and Data Plane Guide, 7.1.0.

  • Hypervisor Details: Verify the hypervisor configuration before deployment.

    Viewing the Hypervisor Details

  • CPU Type: Select Intel Xeon E55xx or newer (Xeon E5‑2xxx recommended). In the Linux shell of the KVM host, you can check the CPU model in /proc/cpuinfo using one of the following commands:
    • grep -m1 '^model name' /proc/cpuinfo
    • lscpu | grep -i '^Model name'

    Selecting the CPU

    Selecting the CPU

  • Memory Allocation: Allocate at least 6 GB (6144 MB) of memory for vSZ-D.

    Allocating the Memory

    Allocating the Memory

  • Disk Configuration: Select SCSI as the disk bus and qcow2 as the storage format.

    Selecting the Disk Configuration

    Selecting the Disk Configuration

  • Controller VirtIO SCSI: Select VirtIO SCSI as the controller.

    Selecting the Controller

    Selecting the Controller

NIC Configuration for Direct I/O Mode

Use Direct I/O mode to assign a physical NIC directly to the VM, bypassing the virtual switch. This setup reduces latency and improves throughput for data traffic.

Note: Only the data interface needs to be configured to direct PCI passthrough (Direct I/O Mode).

Before adding a PCI device to the KVM hypervisor, you need to complete the following steps:

  1. Enable VT-d in the system BIOS (Intel Virtualization Technology for Directed I/O).
    1. Restart the server and enter the BIOS/UEFI setup (usually by pressing Del, F2, or Esc during boot).
    2. Navigate to the CPU or Advanced settings section.
    3. Locate Intel Virtualization Technology for Directed I/O (VT-d).
    4. Set VT-d to Enabled.
    5. Save changes and reboot the system.
  2. Enable IOMMU by adding the kernel parameter intel_iommu=on. For more information, read How to add kernel boot parameters via GRUB on Linux.
    1. Edit the GRUB configuration: 
      sudo vi /etc/default/grub
    2. Add the IOMMU parameter to the kernel line: 
      GRUB_CMDLINE_LINUX="... intel_iommu=on"
    3. Regenerate the GRUB configuration: 
      sudo grub2-mkconfig -o /boot/grub2/grub.cfg
  3. Reboot the system to apply changes.

Reference Configurations (Non-Validated)

Use the following examples only for lab testing. These distributions are not validated for production.

  1. Debian or Ubuntu:
    1. Edit the GRUB template: 
      sudo vi /etc/default/grub
    2. Add the IOMMU parameter: 
      GRUB_CMDLINE_LINUX_DEFAULT="... intel_iommu=on"
    3. Update GRUB: 
      sudo update-grub
    4. If needed, install GRUB tools: 
      sudo apt-get install grub2-common
  2. Fedora:
    1. Edit the GRUB template: 
      sudo vi /etc/default/grub
    2. Add the IOMMU parameter: 
      GRUB_CMDLINE_LINUX="... intel_iommu=on"
    3. Regenerate GRUB configuration: 
      sudo grub2-mkconfig -o /boot/grub2/grub.cfg

NIC Configuration in vSwitch Mode

Use vSwitch mode to connect virtual NICs to the host network through a software switch. This configuration supports flexible traffic routing and is suitable for most virtualized environments.

    Note:
  • Configure two network interfaces: one for management and one for data traffic.
  • Use the e1000 device model for both interfaces to ensure compatibility and performance.

Management Interface Settings

Setting Value
Device Model e1000
Source Mode Bridge or Passthrough (if using macvtap as the device type)
Note: Use Bridge mode for general connectivity. Use Passthrough mode with macvtap for direct host access.

Data Interface Settings

Setting Value
Device Model e1000
Source Mode Passthrough (required if using macvtap)
Note: Passthrough mode is required to allow UE traffic to flow through the VM NIC when using macvtap.

Assigning a PCI Device to a VM in Virt-Manager

Use PCI passthrough to connect a physical device directly to a guest VM. This configuration improves performance by allowing the VM to access hardware without virtualization overhead.

    Note:
  • Ensure IOMMU is enabled on the host system before assigning PCI devices.
  • Use Virt-Manager with VNC Viewer to configure hardware settings.

  1. Open the guest VM in Virt-Manager using VNC Viewer.
  2. Click Add Hardware and select PCI Host Device.
  3. Choose the desired PCI device from the list.
  4. Click Finish to confirm the selection.
  5. Power on the guest VM. The assigned PCI device will appear in the guest system.
Parent topic: Hypervisor Configuration