Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

Guide That Contains This Content
[+] Expand All
[-] Collapse All

    Configuring Multi-queue Virtio Interface for vSRX VM with OpenStack

    Before you plan to scale up vSRX performance, enable network multi-queuing as a means to support an increased number of dataplane vCPUs for the vSRX VM. The default for vSRX in Contrail is 2 dataplane vCPUs, but you can scale that number to 4 vCPUs.

    OpenStack Liberty supports the ability to create VMs with multiple queues on their virtio interfaces. Virtio is a Linux platform for I/O virtualization, providing a common set of I/O virtualization drivers. Multiqueue virtio is an approach that enables the processing of packet sending and receiving to be scaled to the number of available virtual CPUs (vCPUs) of a guest, through the use of multiple queues

    To use multiqueue virtio interfaces, ensure your system meets the following requirements:

    • The OpenStack version must be Liberty or greater.
    • The maximum number of queues in the vSRX VM interface is set to the same value as the number of vCPUs in the guest.
    • The vSRX VM image metadata property is set to enable multiple queues inside the VM.

    Use the following command on the OpenStack node to enable multiple queues on a vSRX VM in Contrail:

    source /etc/contrail/openstackrc

    nova image-meta <image_name> set hw_vif_multiqueue_enabled="true"

    After the vSRX VM is spawned, use the following command on the virtio interface in the guest to enable multiple queues inside the vSRX VM:

    ethtool –L <interface_name> combined <#queues>

    Modified: 2018-01-24