Chapter 4. Configure OVS-DPDK Support for Virtual Networking

This section deploys DPDK with Open vSwitch (OVS-DPDK) within the Red Hat OpenStack Platform environment. The overcloud usually consists of nodes in predefined roles such as Controller nodes, Compute nodes, and different storage node types. Each of these default roles contains a set of services defined in the core Heat templates on the director node.

See Planning your OVS-DPDK Deployment for details on how to determine the best values for the OVS-DPDK parameters that you set in the network-environment.yaml file to optimize your OpenStack network for OVS-DPDK.

4.1. Naming Conventions

We recommend that you follow a consistent naming convention when you use custom roles in your OpenStack deployment, especially with multiple nodes. This naming convention can assist you when creating the following files and configurations:

instackenv.json - To differentiate between nodes with different hardware or NIC capabilities.
```
"name":"computeovsdpdk-0"
```
roles_data.yaml - To differentiate between compute-based roles that support DPDK.
```
`ComputeOvsDpdk`
```
network_environment.yaml - To ensure that you match the custom role to the correct flavor name.
```
 `OvercloudComputeOvsDpdkFlavor: computeovsdpdk`
```
nic-config file names - To differentiate NIC yaml files for compute nodes that support DPDK interfaces.

Flavor creation - To help you match a flavor and capabilities:profile value to the appropriate bare metal node and custom role.

# openstack flavor create --id auto --ram 4096 --disk 40 --vcpus 4 computeovsdpdk
# openstack flavor set --property "cpu_arch"="x86_64" --property "capabilities:boot_option"="local" --property "capabilities:profile"="computeovsdpdk" computeovsdpdk

Bare metal node - To ensure that you match the bare metal node with the appropriate hardware and capability:profile value.

 # openstack baremetal node update computeovsdpdk-0 add properties/capabilities='profile:computeovsdpdk,boot_option:local'

Note

The flavor name does not have to match the capabilities:profile value for the flavor, but the flavor capabilities:profile value must match the bare metal node properties/capabilities='profile value. All three use computeovsdpdk in this example.

Note

Ensure that all your nodes used for a custom role and profile have the same CPU, RAM, and PCI hardware topology.

4.2. OVS-DPDK and Composable Roles

With Red Hat OpenStack Platform 11, you can create custom deployment roles, using the composable roles feature, adding or removing services from each role. For more information on Composable Roles, see Composable Roles and Services.

This image shows a sample OVS-DPDK topology with two bonded ports for the control plane and data plane:

OpenStack NFV Config Guide Topology 450694 0617 ECE OVS DPDK

Configuring OVS-DPDK comprises the following tasks:

If you use composable roles, copy and modify the roles-data.yaml file to add the composable role for OVS-DPDK.
Update the appropriate network-environment.yaml file to include parameters for kernel arguments and DPDK arguments.
Update the compute.yaml file to include the bridge for DPDK interface parameters.
Update the controller.yaml file to include the same bridge details for DPDK interface parameters.
Run the overcloud_deploy.sh script to deploy the overcloud with the DPDK parameters.

Note

This guide provides examples for CPU assignments, memory allocation, and NIC configurations that may vary from your topology and use case. See the Network Functions Virtualization Product Guide and the Network Functions Virtualization Planning and Prerequisite Guide to understand the hardware and configuration options.

Before you begin the procedure, ensure that you have the following:

Red Hat OpenStack Platform 11 with Red Hat Enterprise Linux 7.3
OVS 2.6
DPDK 16.04
Tested NIC. For a list of tested NICs for NFV, see Tested NICs.

Note

Red Hat OpenStack Platform 11 operates in OVS client mode for OVS-DPDK deployments.

4.3. Configure Single-Port OVS-DPDK with VLAN Tunnelling

This section describes how to configure OVS-DPDK with one data plane port.

4.3.1. Modify first-boot.yaml

Modify the first-boot.yaml file to set up OVS and DPDK parameters and to configure tuned for CPU affinity.

Add additional resources.

  resources:
    userdata:
      type: OS::Heat::MultipartMime
      properties:
        parts:
        - config: {get_resource: boot_config}
        - config: {get_resource: set_ovs_socket_config}
        - config: {get_resource: set_ovs_config}
        - config: {get_resource: set_dpdk_params}
        - config: {get_resource: install_tuned}
        - config: {get_resource: compute_kernel_args}

Determine the NeutronVhostUserSocketDir setting.

  set_ovs_socket_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              mkdir -p $NEUTRON_VHOSTUSER_SOCKET_DIR
              chown -R qemu:qemu $NEUTRON_VHOSTUSER_SOCKET_DIR
              restorecon $NEUTRON_VHOSTUSER_SOCKET_DIR
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $NEUTRON_VHOSTUSER_SOCKET_DIR: {get_param: NeutronVhostuserSocketDir}

Set the OVS configuration.

  set_ovs_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              if [ -f /usr/lib/systemd/system/openvswitch-nonetwork.service ]; then
                ovs_service_path="/usr/lib/systemd/system/openvswitch-nonetwork.service"
              elif [ -f /usr/lib/systemd/system/ovs-vswitchd.service ]; then
                ovs_service_path="/usr/lib/systemd/system/ovs-vswitchd.service"
              fi
              grep -q "RuntimeDirectoryMode=.*" $ovs_service_path
              if [ "$?" -eq 0 ]; then
                sed -i 's/RuntimeDirectoryMode=.*/RuntimeDirectoryMode=0775/' $ovs_service_path
              else
                echo "RuntimeDirectoryMode=0775" >> $ovs_service_path
              fi
              grep -Fxq "Group=qemu" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "Group=qemu" >> $ovs_service_path
              fi
              grep -Fxq "UMask=0002" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "UMask=0002" >> $ovs_service_path
              fi
              ovs_ctl_path='/usr/share/openvswitch/scripts/ovs-ctl'
              grep -q "umask 0002 \&\& start_daemon \"\$OVS_VSWITCHD_PRIORITY\"" $ovs_ctl_path
              if [ ! "$?" -eq 0 ]; then
                sed -i 's/start_daemon \"\$OVS_VSWITCHD_PRIORITY.*/umask 0002 \&\& start_daemon \"$OVS_VSWITCHD_PRIORITY\" \"$OVS_VSWITCHD_WRAPPER\" \"$@\"/' $ovs_ctl_path
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}

Set the DPDK parameters.

  set_dpdk_params:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            get_mask()
            {
              local list=$1
              local mask=0
              declare -a bm
              max_idx=0
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  bm[$index]=0
                  if [ $max_idx -lt $index ]; then
                     max_idx=$(($index))
                  fi
              done
              for ((i=$max_idx;i>=0;i--));
              do
                  bm[$i]=0
              done
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  temp=$((1<<$(($core % 32))))
                  bm[$index]=$((${bm[$index]} | $temp))
              done

              printf -v mask "%x" "${bm[$max_idx]}"
              for ((i=$max_idx-1;i>=0;i--));
              do
                  printf -v hex "%08x" "${bm[$i]}"
                  mask+=$hex
              done
              printf "%s" "$mask"
            }

            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              pmd_cpu_mask=$( get_mask $PMD_CORES )
              host_cpu_mask=$( get_mask $LCORE_LIST )
              socket_mem=$(echo $SOCKET_MEMORY | sed s/\'//g )
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-init=true
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-socket-mem=$socket_mem
              ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}
            $SOCKET_MEMORY: {get_param: NeutronDpdkSocketMemory}

Set the tuned configuration to enable CPU affinity.

  install_tuned:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              tuned_conf_path="/etc/tuned/cpu-partitioning-variables.conf"
              if [ -n "$TUNED_CORES" ]; then
                grep -q "^isolated_cores" $tuned_conf_path
                if [ "$?" -eq 0 ]; then
                  sed -i 's/^isolated_cores=.*/isolated_cores=$TUNED_CORES/' $tuned_conf_path
                else
                  echo "isolated_cores=$TUNED_CORES" >> $tuned_conf_path
                fi
                tuned-adm profile cpu-partitioning
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

Set the kernel arguments.

  compute_kernel_args:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              sed 's/^\(GRUB_CMDLINE_LINUX=".*\)"/\1 $KERNEL_ARGS isolcpus=$TUNED_CORES"/g' -i /etc/default/grub ;
              grub2-mkconfig -o /etc/grub2.cfg
              reboot
            fi
          params:
            $KERNEL_ARGS: {get_param: ComputeKernelArgs}
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

4.3.2. Modify post-install.yaml

Set the tuned configuration to enable CPU affinity.

  ExtraConfig:
    type: OS::Heat::SoftwareConfig
    properties:
      group: script
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            function tuned_service_dependency() {
              tuned_service=/usr/lib/systemd/system/tuned.service
              grep -q "network.target" $tuned_service
              if [ "$?" -eq 0 ]; then
                  sed -i '/After=.*/s/network.target//g' $tuned_service
              fi
              grep -q "Before=.*network.target" $tuned_service
              if [ ! "$?" -eq 0 ]; then
                  grep -q "Before=.*" $tuned_service
                  if [ "$?" -eq 0 ]; then
                      sed -i 's/^\(Before=.*\)/\1 network.target openvswitch.service/g' $tuned_service
                  else
                      sed -i '/After/i Before=network.target openvswitch.service' $tuned_service
                  fi
              fi
            }

            get_mask()
            {
                local list=$1
                local mask=0
                declare -a bm
                max_idx=0
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    bm[$index]=0
                    if [ $max_idx -lt $index ]; then
                         max_idx=$(($index))
                    fi
                done
                for ((i=$max_idx;i>=0;i--));
                do
                    bm[$i]=0
                done
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    temp=$((1<<$(($core % 32))))
                    bm[$index]=$((${bm[$index]} | $temp))
                done
                printf -v mask "%x" "${bm[$max_idx]}"
                for ((i=$max_idx-1;i>=0;i--));
                do
                     printf -v hex "%08x" "${bm[$i]}"
                     mask+=$hex
                done
                printf "%s" "$mask"
            }

            if hiera -c /etc/puppet/hiera.yaml service_names | grep -q neutron_ovs_dpdk_agent; then
                pmd_cpu_mask=$( get_mask $PMD_CORES )
                host_cpu_mask=$( get_mask $LCORE_LIST )
                ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
                ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
                tuned_service_dependency
            fi
          params:
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}

4.3.3. Modify network-environment.yaml

Add the custom resources for OVS-DPDK under resource_registry.

resource_registry:
  # Specify the relative/absolute path to the config files you want to use for override the default.
  OS::TripleO::Compute::Net::SoftwareConfig: nic-configs/compute-ovs-dpdk.yaml
  OS::TripleO::Controller::Net::SoftwareConfig: nic-configs/controller.yaml
  OS::TripleO::NodeUserData: first-boot.yaml
  OS::TripleO::NodeExtraConfigPost: post-install.yaml

Under parameter_defaults, disable the tunnel type (set the value to ""), and set the network type to vlan.
```
NeutronTunnelTypes: ''
NeutronNetworkType: 'vlan'
```
Under parameter_defaults, map the physical network to the virtual bridge.
```
NeutronBridgeMappings: 'dpdk_data:br-link0'
```
Under parameter_defaults, set the OpenStack Networking ML2 and Open vSwitch VLAN mapping range.
```
NeutronNetworkVLANRanges: 'dpdk_data:22:22'
```
This example sets the VLAN ranges on the physical network (dpdk_data).
Under parameter_defaults, set the OVS-DPDK configuration parameters.
Note
NeutronDPDKCoreList and NeutronDPDKMemoryChannels are the required settings for this procedure. Attempting to deploy DPDK without appropriate values causes the deployment to fail or lead to unstable deployments.
1. Provide a list of cores that can be used as DPDK poll mode drivers (PMDs) in the format - [allowed_pattern: "[0-9,-]+"].
```
NeutronDpdkCoreList: "1,17,9,25"
```
  Note
  You must assign at least one CPU (with sibling thread) on each NUMA node with or without DPDK NICs present for DPDK PMD to avoid failures in creating guest instances.
To optimize OVS-DPDK performance, consider the following options:
- Select CPUs associated with the NUMA node of the DPDK interface. Use cat /sys/class/net/<interface>/device/numa_node to list the NUMA node associated with an interface and use lscpu to list the CPUs associated with that NUMA node.
- Group CPU siblings together (in case of hyper-threading). Use cat /sys/devices/system/cpu/<cpu>/topology/thread_siblings_list to find the sibling of a CPU.
- Reserve CPU 0 for the host process.
- Isolate CPUs assigned to PMD so that the host process does not use these CPUs.
- Use NovaVcpuPinset to exclude CPUs assigned to PMD from Compute scheduling.
  1. Provide the number of memory channels in the format - [allowed_pattern: "[0-9]+"].
    NeutronDpdkMemoryChannels: "4"
  2. Set the memory pre-allocated from the hugepage pool for each socket.
    NeutronDpdkSocketMemory: "2048,2048"
    This is a comma-separated string, in ascending order of the CPU socket. This example assumes a 2 NUMA node configuration and pre-allocates 2048 MB of huge pages to socket 0, and pre-allocates 2048 MB to socket 0. If you have a single NUMA node system, set this value to 1024,0.
  3. Set the DPDK driver type and the datapath type for OVS bridges.
    NeutronDpdkDriverType: "vfio-pci" NeutronDatapathType: "netdev"
Under parameter_defaults, set the vhost-user socket directory for OVS.
```
NeutronVhostuserSocketDir: "/var/lib/vhost_sockets"
```
Under parameter_defaults, reserve the RAM for the host processes.
```
NovaReservedHostMemory: 2048
```
Under parameter_defaults, set a comma-separated list or range of physical CPU cores to reserve for virtual machine processes.
```
NovaVcpuPinSet: "2,3,4,5,6,7,18,19,20,21,22,23,10,11,12,13,14,15,26,27,28,29,30,31"
```
Under parameter_defaults, list the applicable filters.
Nova scheduler applies these filters in the order they are listed. List the most restrictive filters first to make the filtering process for the nodes more efficient.
```
NovaSchedulerDefaultFilters: "RamFilter,ComputeFilter,AvailabilityZoneFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,PciPassthroughFilter,NUMATopologyFilter"
```
Under parameter_defaults, define the ComputeKernelArgs parameters to be included in the default grub file at first boot.
```
ComputeKernelArgs: "default_hugepagesz=1GB hugepagesz=1G hugepages=32 iommu=pt intel_iommu=on"
```
Note
These huge pages are consumed by the virtual machines, and also by OVS-DPDK using the NeutronDpdkSocketMemory parameter as shown in this procedure. The number of huge pages available for the virtual machines is the boot parameter minus the NeutronDpdkSocketMemory.
You need to add hw:mem_page_size=1GB to the flavor you associate with the DPDK instance. If you do not do this, the instance does not get a DHCP allocation.
Under parameter_defaults, set a list or range of physical CPU cores to be isolated from the host.
The given argument is appended to the tuned cpu-partitioning profile.
```
HostIsolatedCoreList: "1,2,3,4,5,6,7,9,10,17,18,19,20,21,22,23,11,12,13,14,15,25,26,27,28,29,30,31"
```
Under parameter_defaults, set a list of logical cores used by OVS-DPDK processes for dpdk-lcore-mask. These cores must be mutually exclusive from the list of cores in NeutronDpdkCoreList and NovaVcpuPinSet.
```
HostCpusList: "0,16,8,24"
```

4.3.4. Modify controller.yaml

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageMgmtNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageMgmtIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: ExternalNetworkVlanID}
    device: bond_api
    addresses:
    -
      ip_netmask: {get_param: ExternalIpSubnet}
    routes:
      -
        default: true
        next_hop: {get_param: ExternalInterfaceDefaultRoute}

Create the OVS bridge to the Compute node.

  -
    type: ovs_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic4

4.3.5. Modify compute.yaml

Create the compute-ovs-dpdk.yaml file from the default compute.yaml file and make the following changes:

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}

Set a bridge with a DPDK port to link to the controller.
```
  -
    type: ovs_user_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk0
        members:
          -
            type: interface
            name: nic4
```
Note
To include multiple DPDK devices, repeat the type code section for each DPDK device you want to add.
Note
When using OVS-DPDK, all bridges on the same Compute node should be of type ovs_user_bridge. The director may accept the configuration, but Red Hat OpenStack Platform does not support mixing ovs_bridge and ovs_user_bridge on the same node.

4.3.6. Run the overcloud_deploy.sh Script

The following example defines the openstack overcloud deploy command for the OVS-DPDK environment within a bash script:

#!/bin/bash

openstack overcloud deploy \
--templates \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml \
-e /home/stack/ospd-11-vlan-dpdk-single-port-ctlplane-bonding/network-environment.yaml

/usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml is the location of the default neutron-ovs-dpdk.yaml file, which enables the OVS-DPDK parameters for the Compute role.
/home/stack/<relative-directory>/network-environment.yaml is the path for the network-environment.yaml file. Use this file to overwrite the default values from the neutron-ovs-dpdk.yaml file.

Note

Reboot the Compute nodes to enforce the tuned profile after the overcloud is deployed.

Note

This configuration of OVS-DPDK does not support security groups and live migrations.

4.4. Configure Two-Port OVS-DPDK with VLAN Tunnelling

This section describes how to configure OVS-DPDK with two data plane ports.

4.4.1. Modify first-boot.yaml

Modify the first-boot.yaml file to set up OVS and DPDK parameters and to configure tuned for CPU affinity.

Add additional resources.

resources:
  userdata:
    type: OS::Heat::MultipartMime
    properties:
      parts:
      - config: {get_resource: boot_config}
      - config: {get_resource: set_ovs_socket_config}
      - config: {get_resource: set_ovs_config}
      - config: {get_resource: set_dpdk_params}
      - config: {get_resource: install_tuned}
      - config: {get_resource: compute_kernel_args}

Determine the NeutronVhostUserSocketDir setting.

  set_ovs_socket_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              mkdir -p $NEUTRON_VHOSTUSER_SOCKET_DIR
              chown -R qemu:qemu $NEUTRON_VHOSTUSER_SOCKET_DIR
              restorecon $NEUTRON_VHOSTUSER_SOCKET_DIR
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $NEUTRON_VHOSTUSER_SOCKET_DIR: {get_param: NeutronVhostuserSocketDir}

Set the OVS configuration.

  set_ovs_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              if [ -f /usr/lib/systemd/system/openvswitch-nonetwork.service ]; then
                ovs_service_path="/usr/lib/systemd/system/openvswitch-nonetwork.service"
              elif [ -f /usr/lib/systemd/system/ovs-vswitchd.service ]; then
                ovs_service_path="/usr/lib/systemd/system/ovs-vswitchd.service"
              fi
              grep -q "RuntimeDirectoryMode=.*" $ovs_service_path
              if [ "$?" -eq 0 ]; then
                sed -i 's/RuntimeDirectoryMode=.*/RuntimeDirectoryMode=0775/' $ovs_service_path
              else
                echo "RuntimeDirectoryMode=0775" >> $ovs_service_path
              fi
              grep -Fxq "Group=qemu" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "Group=qemu" >> $ovs_service_path
              fi
              grep -Fxq "UMask=0002" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "UMask=0002" >> $ovs_service_path
              fi
              ovs_ctl_path='/usr/share/openvswitch/scripts/ovs-ctl'
              grep -q "umask 0002 \&\& start_daemon \"\$OVS_VSWITCHD_PRIORITY\"" $ovs_ctl_path
              if [ ! "$?" -eq 0 ]; then
                sed -i 's/start_daemon \"\$OVS_VSWITCHD_PRIORITY.*/umask 0002 \&\& start_daemon \"$OVS_VSWITCHD_PRIORITY\" \"$OVS_VSWITCHD_WRAPPER\" \"$@\"/' $ovs_ctl_path
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}

Set the DPDK parameters.

  set_dpdk_params:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            get_mask()
            {
              local list=$1
              local mask=0
              declare -a bm
              max_idx=0
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  bm[$index]=0
                  if [ $max_idx -lt $index ]; then
                     max_idx=$(($index))
                  fi
              done
              for ((i=$max_idx;i>=0;i--));
              do
                  bm[$i]=0
              done
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  temp=$((1<<$(($core % 32))))
                  bm[$index]=$((${bm[$index]} | $temp))
              done

              printf -v mask "%x" "${bm[$max_idx]}"
              for ((i=$max_idx-1;i>=0;i--));
              do
                  printf -v hex "%08x" "${bm[$i]}"
                  mask+=$hex
              done
              printf "%s" "$mask"
            }

            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              pmd_cpu_mask=$( get_mask $PMD_CORES )
              host_cpu_mask=$( get_mask $LCORE_LIST )
              socket_mem=$(echo $SOCKET_MEMORY | sed s/\'//g )
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-init=true
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-socket-mem=$socket_mem
              ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}
            $SOCKET_MEMORY: {get_param: NeutronDpdkSocketMemory}

Set the tuned configuration to enable CPU affinity.

  install_tuned:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              tuned_conf_path="/etc/tuned/cpu-partitioning-variables.conf"
              if [ -n "$TUNED_CORES" ]; then
                grep -q "^isolated_cores" $tuned_conf_path
                if [ "$?" -eq 0 ]; then
                  sed -i 's/^isolated_cores=.*/isolated_cores=$TUNED_CORES/' $tuned_conf_path
                else
                  echo "isolated_cores=$TUNED_CORES" >> $tuned_conf_path
                fi
                tuned-adm profile cpu-partitioning
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

Set the kernel arguments.

  compute_kernel_args:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              sed 's/^\(GRUB_CMDLINE_LINUX=".*\)"/\1 $KERNEL_ARGS isolcpus=$TUNED_CORES"/g' -i /etc/default/grub ;
              grub2-mkconfig -o /etc/grub2.cfg
              reboot
            fi
          params:
            $KERNEL_ARGS: {get_param: ComputeKernelArgs}
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

4.4.2. Modify post-install.yaml

Set the tuned configuration to enable CPU affinity.

  ExtraConfig:
    type: OS::Heat::SoftwareConfig
    properties:
      group: script
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            function tuned_service_dependency() {
              tuned_service=/usr/lib/systemd/system/tuned.service
              grep -q "network.target" $tuned_service
              if [ "$?" -eq 0 ]; then
                  sed -i '/After=.*/s/network.target//g' $tuned_service
              fi
              grep -q "Before=.*network.target" $tuned_service
              if [ ! "$?" -eq 0 ]; then
                  grep -q "Before=.*" $tuned_service
                  if [ "$?" -eq 0 ]; then
                      sed -i 's/^\(Before=.*\)/\1 network.target openvswitch.service/g' $tuned_service
                  else
                      sed -i '/After/i Before=network.target openvswitch.service' $tuned_service
                  fi
              fi
            }

            get_mask()
            {
                local list=$1
                local mask=0
                declare -a bm
                max_idx=0
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    bm[$index]=0
                    if [ $max_idx -lt $index ]; then
                         max_idx=$(($index))
                    fi
                done
                for ((i=$max_idx;i>=0;i--));
                do
                    bm[$i]=0
                done
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    temp=$((1<<$(($core % 32))))
                    bm[$index]=$((${bm[$index]} | $temp))
                done
                printf -v mask "%x" "${bm[$max_idx]}"
                for ((i=$max_idx-1;i>=0;i--));
                do
                     printf -v hex "%08x" "${bm[$i]}"
                     mask+=$hex
                done
                printf "%s" "$mask"
            }

            if hiera -c /etc/puppet/hiera.yaml service_names | grep -q neutron_ovs_dpdk_agent; then
                pmd_cpu_mask=$( get_mask $PMD_CORES )
                host_cpu_mask=$( get_mask $LCORE_LIST )
                ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
                ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
                tuned_service_dependency
            fi
          params:
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}

4.4.3. Modify network-environment.yaml

Add the custom resources for OVS-DPDK under resource_registry.

resource_registry:
  # Specify the relative/absolute path to the config files you want to use for override the default.
  OS::TripleO::Compute::Net::SoftwareConfig: nic-configs/compute-ovs-dpdk.yaml
  OS::TripleO::Controller::Net::SoftwareConfig: nic-configs/controller.yaml
  OS::TripleO::NodeUserData: first-boot.yaml
  OS::TripleO::NodeExtraConfigPost: post-install.yaml

Under parameter_defaults, disable the tunnel type (set the value to ""), and set the network type to vlan.
```
NeutronTunnelTypes: ''
NeutronNetworkType: 'vlan'
```
Under parameter_defaults, map the physical network to the virtual bridge.
```
NeutronBridgeMappings: 'dpdk_mgmt:br-link0,dpdk_data:br-link1'
```
Under parameter_defaults, set the OpenStack Networking ML2 and Open vSwitch VLAN mapping range.
```
NeutronNetworkVLANRanges: 'dpdk_mgmt:22:22,dpdk_data:25:28'
```
Under parameter_defaults, set the OVS-DPDK configuration parameters.
Note
NeutronDPDKCoreList and NeutronDPDKMemoryChannels are the required settings for this procedure. Attempting to deploy DPDK without appropriate values causes the deployment to fail or lead to unstable deployments.
1. Provide a list of cores that can be used as DPDK poll mode drivers (PMDs) in the format - [allowed_pattern: "[0-9,-]+"].
```
NeutronDpdkCoreList: "1,17,9,25"
```
  Note
  You must assign at least one CPU (with sibling thread) on each NUMA node with or without DPDK NICs present for DPDK PMD to avoid failures in creating guest instances.
To optimize OVS-DPDK performance, consider the following options:
- Select CPUs associated with the NUMA node of the DPDK interface. Use cat /sys/class/net/<interface>/device/numa_node to list the NUMA node associated with an interface and use lscpu to list the CPUs associated with that NUMA node.
- Group CPU siblings together (in case of hyper-threading). Use cat /sys/devices/system/cpu/<cpu>/topology/thread_siblings_list to find the sibling of a CPU.
- Reserve CPU 0 for the host process.
- Isolate CPUs assigned to PMD so that the host process does not use these CPUs.
- Use NovaVcpuPinset to exclude CPUs assigned to PMD from Compute scheduling.
  1. Provide the number of memory channels in the format - [allowed_pattern: "[0-9]+"].
    NeutronDpdkMemoryChannels: "4"
  2. Set the memory pre-allocated from the hugepage pool for each socket.
    NeutronDpdkSocketMemory: "2048,2048"
    This is a comma-separated string, in ascending order of the CPU socket. This example assumes a 2 NUMA node configuration and pre-allocates 2048 MB of huge pages to socket 0, and pre-allocates 2048 MB to socket 0. If you have a single NUMA node system, set this value to 1024,0.
  3. Set the DPDK driver type and the datapath type for OVS bridges.
    NeutronDpdkDriverType: "vfio-pci" NeutronDatapathType: "netdev"
Under parameter_defaults, set the vhost-user socket directory for OVS.
```
NeutronVhostuserSocketDir: "/var/lib/vhost_sockets"
```
Under parameter_defaults, reserve the RAM for the host processes.
```
NovaReservedHostMemory: 2048
```
Under parameter_defaults, set a comma-separated list or range of physical CPU cores to reserve for virtual machine processes.
```
NovaVcpuPinSet: "2,3,4,5,6,7,18,19,20,21,22,23,10,11,12,13,14,15,26,27,28,29,30,31"
```
Under parameter_defaults, list the applicable filters.
Nova scheduler applies these filters in the order they are listed. List the most restrictive filters first to make the filtering process for the nodes more efficient.
```
NovaSchedulerDefaultFilters: "RamFilter,ComputeFilter,AvailabilityZoneFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,PciPassthroughFilter,NUMATopologyFilter"
```
Under parameter_defaults, define the ComputeKernelArgs parameters to be included in the default grub file at first boot.
```
ComputeKernelArgs: "default_hugepagesz=1GB hugepagesz=1G hugepages=32 iommu=pt intel_iommu=on"
```
Note
These huge pages are consumed by the virtual machines, and also by OVS-DPDK using the NeutronDpdkSocketMemory parameter as shown in this procedure. The number of huge pages available for the virtual machines is the boot parameter minus the NeutronDpdkSocketMemory.
You need to add hw:mem_page_size=1GB to the flavor you associate with the DPDK instance. If you do not do this, the instance does get a DHCP allocation.
Under parameter_defaults, set a list or range of physical CPU cores to be isolated from the host.
The given argument is appended to the tuned cpu-partitioning profile.
```
HostIsolatedCoreList: "1,2,3,4,5,6,7,9,10,17,18,19,20,21,22,23,11,12,13,14,15,25,26,27,28,29,30,31"
```
Under parameter_defaults, set a list of logical cores used by OVS-DPDK processes for dpdk-lcore-mask. These cores must be mutually exclusive from the list of cores in NeutronDpdkCoreList and NovaVcpuPinSet.
```
HostCpusList: "0,16,8,24"
```

4.4.4. Modify controller.yaml

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageMgmtNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageMgmtIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: ExternalNetworkVlanID}
    device: bond_api
    addresses:
    -
      ip_netmask: {get_param: ExternalIpSubnet}
    routes:
      -
        default: true
        next_hop: {get_param: ExternalInterfaceDefaultRoute}

Create two OVS bridges to the Compute node.

  -
    type: ovs_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic4
  -
    type: ovs_bridge
    name: br-link1
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic5

4.4.5. Modify compute.yaml

Create the compute-ovs-dpdk.yaml file from the default compute.yaml file and make the following changes:

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}

Set two bridges with a DPDK port each to link to the controller.

  -
    type: ovs_user_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk0
        members:
          -
            type: interface
            name: nic4
  -
    type: ovs_user_bridge
    name: br-link1
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk1
        members:
          -
            type: interface
            name: nic5

Note

To include multiple DPDK devices, repeat the type code section for each DPDK device you want to add.

Note

When using OVS-DPDK, all bridges on the same Compute node should be of type ovs_user_bridge. The director may accept the configuration, but Red Hat OpenStack Platform does not support mixing ovs_bridge and ovs_user_bridge on the same node.

4.4.6. Run the overcloud_deploy.sh Script

The following example defines the openstack overcloud deploy command for the OVS-DPDK environment within a bash script:

#!/bin/bash

openstack overcloud deploy \
--templates \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml \
-e /home/stack/ospd-11-vlan-dpdk-two-ports-ctlplane-bonding/network-environment.yaml

/usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml is the location of the default neutron-ovs-dpdk.yaml file, which enables the OVS-DPDK parameters for the Compute role.
/home/stack/<relative-directory>/network-environment.yaml is the path for the network-environment.yaml file. Use this file to overwrite the default values from the neutron-ovs-dpdk.yaml file.

Note

Reboot the Compute nodes to enforce the tuned profile after the overcloud is deployed.

Note

This configuration of OVS-DPDK does not support security groups and live migrations.

4.5. Configure Two-Port OVS-DPDK Data Plane Bonding with VLAN Tunnelling

This section describes how to configure OVS-DPDK with two data plane ports in an OVS-DPDK bond.

4.5.1. Modify first-boot.yaml

Modify the first-boot.yaml file to set up OVS and DPDK parameters and to configure tuned for CPU affinity.

Add additional resources.

resources:
  userdata:
    type: OS::Heat::MultipartMime
    properties:
      parts:
      - config: {get_resource: boot_config}
      - config: {get_resource: set_ovs_socket_config}
      - config: {get_resource: set_ovs_config}
      - config: {get_resource: set_dpdk_params}
      - config: {get_resource: install_tuned}
      - config: {get_resource: compute_kernel_args}

Determine the NeutronVhostUserSocketDir setting.

  set_ovs_socket_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              mkdir -p $NEUTRON_VHOSTUSER_SOCKET_DIR
              chown -R qemu:qemu $NEUTRON_VHOSTUSER_SOCKET_DIR
              restorecon $NEUTRON_VHOSTUSER_SOCKET_DIR
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $NEUTRON_VHOSTUSER_SOCKET_DIR: {get_param: NeutronVhostuserSocketDir}

Set the OVS configuration.

  set_ovs_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              if [ -f /usr/lib/systemd/system/openvswitch-nonetwork.service ]; then
                ovs_service_path="/usr/lib/systemd/system/openvswitch-nonetwork.service"
              elif [ -f /usr/lib/systemd/system/ovs-vswitchd.service ]; then
                ovs_service_path="/usr/lib/systemd/system/ovs-vswitchd.service"
              fi
              grep -q "RuntimeDirectoryMode=.*" $ovs_service_path
              if [ "$?" -eq 0 ]; then
                sed -i 's/RuntimeDirectoryMode=.*/RuntimeDirectoryMode=0775/' $ovs_service_path
              else
                echo "RuntimeDirectoryMode=0775" >> $ovs_service_path
              fi
              grep -Fxq "Group=qemu" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "Group=qemu" >> $ovs_service_path
              fi
              grep -Fxq "UMask=0002" $ovs_service_path
              if [ ! "$?" -eq 0 ]; then
                echo "UMask=0002" >> $ovs_service_path
              fi
              ovs_ctl_path='/usr/share/openvswitch/scripts/ovs-ctl'
              grep -q "umask 0002 \&\& start_daemon \"\$OVS_VSWITCHD_PRIORITY\"" $ovs_ctl_path
              if [ ! "$?" -eq 0 ]; then
                sed -i 's/start_daemon \"\$OVS_VSWITCHD_PRIORITY.*/umask 0002 \&\& start_daemon \"$OVS_VSWITCHD_PRIORITY\" \"$OVS_VSWITCHD_WRAPPER\" \"$@\"/' $ovs_ctl_path
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}

Set the DPDK parameters.

  set_dpdk_params:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            get_mask()
            {
              local list=$1
              local mask=0
              declare -a bm
              max_idx=0
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  bm[$index]=0
                  if [ $max_idx -lt $index ]; then
                     max_idx=$(($index))
                  fi
              done
              for ((i=$max_idx;i>=0;i--));
              do
                  bm[$i]=0
              done
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  temp=$((1<<$(($core % 32))))
                  bm[$index]=$((${bm[$index]} | $temp))
              done

              printf -v mask "%x" "${bm[$max_idx]}"
              for ((i=$max_idx-1;i>=0;i--));
              do
                  printf -v hex "%08x" "${bm[$i]}"
                  mask+=$hex
              done
              printf "%s" "$mask"
            }

            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              pmd_cpu_mask=$( get_mask $PMD_CORES )
              host_cpu_mask=$( get_mask $LCORE_LIST )
              socket_mem=$(echo $SOCKET_MEMORY | sed s/\'//g )
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-init=true
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-socket-mem=$socket_mem
              ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}
            $SOCKET_MEMORY: {get_param: NeutronDpdkSocketMemory}

Set the tuned configuration to enable CPU affinity.

  install_tuned:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              tuned_conf_path="/etc/tuned/cpu-partitioning-variables.conf"
              if [ -n "$TUNED_CORES" ]; then
                grep -q "^isolated_cores" $tuned_conf_path
                if [ "$?" -eq 0 ]; then
                  sed -i 's/^isolated_cores=.*/isolated_cores=$TUNED_CORES/' $tuned_conf_path
                else
                  echo "isolated_cores=$TUNED_CORES" >> $tuned_conf_path
                fi
                tuned-adm profile cpu-partitioning
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

Set the kernel arguments.

  compute_kernel_args:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              sed 's/^\(GRUB_CMDLINE_LINUX=".*\)"/\1 $KERNEL_ARGS isolcpus=$TUNED_CORES"/g' -i /etc/default/grub ;
              grub2-mkconfig -o /etc/grub2.cfg
              reboot
            fi
          params:
            $KERNEL_ARGS: {get_param: ComputeKernelArgs}
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

4.5.2. Modify post-install.yaml

Set the tuned configuration to enable CPU affinity.

  ExtraConfig:
    type: OS::Heat::SoftwareConfig
    properties:
      group: script
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            function tuned_service_dependency() {
              tuned_service=/usr/lib/systemd/system/tuned.service
              grep -q "network.target" $tuned_service
              if [ "$?" -eq 0 ]; then
                  sed -i '/After=.*/s/network.target//g' $tuned_service
              fi
              grep -q "Before=.*network.target" $tuned_service
              if [ ! "$?" -eq 0 ]; then
                  grep -q "Before=.*" $tuned_service
                  if [ "$?" -eq 0 ]; then
                      sed -i 's/^\(Before=.*\)/\1 network.target openvswitch.service/g' $tuned_service
                  else
                      sed -i '/After/i Before=network.target openvswitch.service' $tuned_service
                  fi
              fi
            }

            get_mask()
            {
                local list=$1
                local mask=0
                declare -a bm
                max_idx=0
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    bm[$index]=0
                    if [ $max_idx -lt $index ]; then
                         max_idx=$(($index))
                    fi
                done
                for ((i=$max_idx;i>=0;i--));
                do
                    bm[$i]=0
                done
                for core in $(echo $list | sed 's/,/ /g')
                do
                    index=$(($core/32))
                    temp=$((1<<$(($core % 32))))
                    bm[$index]=$((${bm[$index]} | $temp))
                done
                printf -v mask "%x" "${bm[$max_idx]}"
                for ((i=$max_idx-1;i>=0;i--));
                do
                     printf -v hex "%08x" "${bm[$i]}"
                     mask+=$hex
                done
                printf "%s" "$mask"
            }

            if hiera -c /etc/puppet/hiera.yaml service_names | grep -q neutron_ovs_dpdk_agent; then
                pmd_cpu_mask=$( get_mask $PMD_CORES )
                host_cpu_mask=$( get_mask $LCORE_LIST )
                ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
                ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
                tuned_service_dependency
            fi
          params:
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}

4.5.3. Modify network-environment.yaml

Add the custom resources for OVS-DPDK under resource_registry.

resource_registry:
  # Specify the relative/absolute path to the config files you want to use for override the default.
  OS::TripleO::Compute::Net::SoftwareConfig: nic-configs/compute.yaml
  OS::TripleO::Controller::Net::SoftwareConfig: nic-configs/controller.yaml
  OS::TripleO::NodeUserData: first-boot.yaml
  OS::TripleO::NodeExtraConfigPost: post-install.yaml

Under parameter_defaults, disable the tunnel type (set the value to ""), and set the network type to vlan.
```
NeutronTunnelTypes: ''
NeutronNetworkType: 'vlan'
```
Under parameter_defaults, map the physical network to the virtual bridge.
```
NeutronBridgeMappings: 'dpdk_mgmt:br-link'
```
Under parameter_defaults, set the OpenStack Networking ML2 and Open vSwitch VLAN mapping range.
```
NeutronNetworkVLANRanges: 'dpdk_mgmt:22:22'
```
This example sets the VLAN ranges on the physical network (dpdk_mgmt).
Under parameter_defaults, set the OVS-DPDK configuration parameters.
Note
NeutronDPDKCoreList and NeutronDPDKMemoryChannels are the required settings for this procedure. Attempting to deploy DPDK without appropriate values causes the deployment to fail or lead to unstable deployments.
1. Provide a list of cores that can be used as DPDK poll mode drivers (PMDs) in the format - [allowed_pattern: "[0-9,-]+"].
```
NeutronDpdkCoreList: "1,17,9,25"
```
  Note
  You must assign at least one CPU (with sibling thread) on each NUMA node with or without DPDK NICs present for DPDK PMD to avoid failures in creating guest instances.
To optimize OVS-DPDK performance, consider the following options:
- Select CPUs associated with the NUMA node of the DPDK interface. Use cat /sys/class/net/<interface>/device/numa_node to list the NUMA node associated with an interface and use lscpu to list the CPUs associated with that NUMA node.
- Group CPU siblings together (in case of hyper-threading). Use cat /sys/devices/system/cpu/<cpu>/topology/thread_siblings_list to find the sibling of a CPU.
- Reserve CPU 0 for the host process.
- Isolate CPUs assigned to PMD so that the host process does not use these CPUs.
- Use NovaVcpuPinset to exclude CPUs assigned to PMD from Compute scheduling.
  1. Provide the number of memory channels in the format - [allowed_pattern: "[0-9]+"].
    NeutronDpdkMemoryChannels: "4"
  2. Set the memory pre-allocated from the hugepage pool for each socket.
    NeutronDpdkSocketMemory: "1024,1024"
    This is a comma-separated string, in ascending order of the CPU socket. This example assumes a 2 NUMA node configuration and pre-allocates 1024 MB of huge pages to socket 0, and pre-allocates 1024 MB to socket 0. If you have a single NUMA node system, set this value to 1024,0.
  3. Set the DPDK driver type and the datapath type for OVS bridges.
    NeutronDpdkDriverType: "vfio-pci" NeutronDatapathType: "netdev"
Under parameter_defaults, set the vhost-user socket directory for OVS.
```
NeutronVhostuserSocketDir: "/var/lib/vhost_sockets"
```
Under parameter_defaults, reserve the RAM for the host processes.
```
NovaReservedHostMemory: 2048
```
Under parameter_defaults, set a comma-separated list or range of physical CPU cores to reserve for virtual machine processes.
```
NovaVcpuPinSet: "2,3,4,5,6,7,18,19,20,21,22,23,10,11,12,13,14,15,26,27,28,29,30,31"
```
Under parameter_defaults, list the applicable filters.
Nova scheduler applies these filters in the order they are listed. List the most restrictive filters first to make the filtering process for the nodes more efficient.
```
NovaSchedulerDefaultFilters: "RamFilter,ComputeFilter,AvailabilityZoneFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,PciPassthroughFilter,NUMATopologyFilter"
```
Under parameter_defaults, define the ComputeKernelArgs parameters to be included in the default grub file at first boot.
```
ComputeKernelArgs: "default_hugepagesz=1GB hugepagesz=1G hugepages=32 iommu=pt intel_iommu=on"
```
Note
These huge pages are consumed by the virtual machines, and also by OVS-DPDK using the NeutronDpdkSocketMemory parameter as shown in this procedure. The number of huge pages available for the virtual machines is the boot parameter minus the NeutronDpdkSocketMemory.
You need to add hw:mem_page_size=1GB to the flavor you associate with the DPDK instance. If you do not do this, the instance does not get a DHCP allocation.
Under parameter_defaults, set a list or range of physical CPU cores to be isolated from the host.
The given argument is appended to the tuned cpu-partitioning profile.
```
HostIsolatedCoreList: "1,2,3,4,5,6,7,9,10,17,18,19,20,21,22,23,11,12,13,14,15,25,26,27,28,29,30,31"
```
Under parameter_defaults, set a list of logical cores used by OVS-DPDK processes for dpdk-lcore-mask. These cores must be mutually exclusive from the list of cores in NeutronDpdkCoreList and NovaVcpuPinSet.
```
HostCpusList: "0,16,8,24"
```

4.5.4. Modify controller.yaml

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond1
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond1
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: ExternalNetworkVlanID}
    device: bond1
    addresses:
      -
        ip_netmask: {get_param: ExternalIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond1
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}

Create the OVS bridge to the Compute node.

  -
    type: ovs_bridge
    name: br-link
    use_dhcp: false
    members:
      -
        type: interface
        name: nic4

4.5.5. Modify compute.yaml

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: TenantNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}

Set a bridge with two DPDK ports in an OVS-DPDK bond to link to the controller.

  -
    type: ovs_user_bridge
    name: br-link
    use_dhcp: false
    members:
      -
        type: ovs_dpdk_bond
        name: bond_dpdk0
        members:
          -
            type: ovs_dpdk_port
            name: dpdk0
            members:
              -
                type: interface
                name: nic4
          -
            type: ovs_dpdk_port
            name: dpdk1
            members:
              -
                type: interface
                name: nic5

Note

To include multiple DPDK devices, repeat the type code section for each DPDK device you want to add.

Note

4.5.6. Run the overcloud_deploy.sh Script

The following example defines the openstack overcloud deploy command for the OVS-DPDK environment within a bash script:

#!/bin/bash

openstack overcloud deploy \
--templates \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml \
-e /home/stack/ospd-11-vlan-ovs-dpdk-bonding-dataplane-bonding-ctlplane/network-environment.yaml

/usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml is the location of the default neutron-ovs-dpdk.yaml file, which enables the OVS-DPDK parameters for the Compute role.
/home/stack/<relative-directory>/network-environment.yaml is the path for the network-environment.yaml file. Use this file to overwrite the default values from the neutron-ovs-dpdk.yaml file.

Note

Reboot the Compute nodes to enforce the tuned profile after the overcloud is deployed.

Note

This configuration of OVS-DPDK does not support security groups and live migrations.

4.6. Configure Single-Port OVS-DPDK with VXLAN Tunnelling

This section describes how to configure OVS-DPDK with VXLAN tunnelling.

4.6.1. Modify first-boot.yaml

Modify the first-boot.yaml file to set up OVS and DPDK parameters and to configure tuned for CPU affinity.

Add additional resources.

resources:
  userdata:
    type: OS::Heat::MultipartMime
    properties:
      parts:
      - config: {get_resource: set_ovs_socket_config}
      - config: {get_resource: set_ovs_config}
      - config: {get_resource: set_dpdk_params}
      - config: {get_resource: install_tuned}
      - config: {get_resource: compute_kernel_args}

Determine the NeutronVhostUserSocketDir setting.

  set_ovs_socket_config:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              mkdir -p $NEUTRON_VHOSTUSER_SOCKET_DIR
              chown -R qemu:qemu $NEUTRON_VHOSTUSER_SOCKET_DIR
              restorecon $NEUTRON_VHOSTUSER_SOCKET_DIR
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $NEUTRON_VHOSTUSER_SOCKET_DIR: {get_param: NeutronVhostuserSocketDir}

Set the OVS configuration.

set_ovs_config:
  type: OS::Heat::SoftwareConfig
  properties:
    config:
      str_replace:
        template: |
          #!/bin/bash
          FORMAT=$COMPUTE_HOSTNAME_FORMAT
          if [[ -z $FORMAT ]] ; then
            FORMAT="compute" ;
          else
            # Assumption: only %index% and %stackname% are the variables in Host name format
            FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
          fi
          if [[ $(hostname) == *$FORMAT* ]] ; then
            if [ -f /usr/lib/systemd/system/openvswitch-nonetwork.service ]; then
              ovs_service_path="/usr/lib/systemd/system/openvswitch-nonetwork.service"
            elif [ -f /usr/lib/systemd/system/ovs-vswitchd.service ]; then
              ovs_service_path="/usr/lib/systemd/system/ovs-vswitchd.service"
            fi
            grep -q "RuntimeDirectoryMode=.*" $ovs_service_path
            if [ "$?" -eq 0 ]; then
              sed -i 's/RuntimeDirectoryMode=.*/RuntimeDirectoryMode=0775/' $ovs_service_path
            else
              echo "RuntimeDirectoryMode=0775" >> $ovs_service_path
            fi
            grep -Fxq "Group=qemu" $ovs_service_path
            if [ ! "$?" -eq 0 ]; then
              echo "Group=qemu" >> $ovs_service_path
            fi
            grep -Fxq "UMask=0002" $ovs_service_path
            if [ ! "$?" -eq 0 ]; then
              echo "UMask=0002" >> $ovs_service_path
            fi
            ovs_ctl_path='/usr/share/openvswitch/scripts/ovs-ctl'
            grep -q "umask 0002 \&\& start_daemon \"\$OVS_VSWITCHD_PRIORITY\"" $ovs_ctl_path
            if [ ! "$?" -eq 0 ]; then
              sed -i 's/start_daemon \"\$OVS_VSWITCHD_PRIORITY.*/umask 0002 \&\& start_daemon \"$OVS_VSWITCHD_PRIORITY\" \"$OVS_VSWITCHD_WRAPPER\" \"$@\"/' $ovs_ctl_path
            fi
          fi
        params:
          $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}

Set the DPDK parameters.

  set_dpdk_params:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            set -x
            get_mask()
            {
              local list=$1
              local mask=0
              declare -a bm
              max_idx=0
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  bm[$index]=0
                  if [ $max_idx -lt $index ]; then
                     max_idx=$(($index))
                  fi
              done
              for ((i=$max_idx;i>=0;i--));
              do
                  bm[$i]=0
              done
              for core in $(echo $list | sed 's/,/ /g')
              do
                  index=$(($core/32))
                  temp=$((1<<$(($core % 32))))
                  bm[$index]=$((${bm[$index]} | $temp))
              done

              printf -v mask "%x" "${bm[$max_idx]}"
              for ((i=$max_idx-1;i>=0;i--));
              do
                  printf -v hex "%08x" "${bm[$i]}"
                  mask+=$hex
              done
              printf "%s" "$mask"
            }

            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              pmd_cpu_mask=$( get_mask $PMD_CORES )
              host_cpu_mask=$( get_mask $LCORE_LIST )
              socket_mem=$(echo $SOCKET_MEMORY | sed s/\'//g )
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-init=true
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-socket-mem=$socket_mem
              ovs-vsctl --no-wait set Open_vSwitch . other_config:pmd-cpu-mask=$pmd_cpu_mask
              ovs-vsctl --no-wait set Open_vSwitch . other_config:dpdk-lcore-mask=$host_cpu_mask
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $LCORE_LIST: {get_param: HostCpusList}
            $PMD_CORES: {get_param: NeutronDpdkCoreList}
            $SOCKET_MEMORY: {get_param: NeutronDpdkSocketMemory}

Set the tuned configuration to enable CPU affinity.

  install_tuned:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              tuned_conf_path="/etc/tuned/cpu-partitioning-variables.conf"
              if [ -n "$TUNED_CORES" ]; then
                grep -q "^isolated_cores" $tuned_conf_path
                if [ "$?" -eq 0 ]; then
                  sed -i 's/^isolated_cores=.*/isolated_cores=$TUNED_CORES/' $tuned_conf_path
                else
                  echo "isolated_cores=$TUNED_CORES" >> $tuned_conf_path
                fi
                tuned-adm profile cpu-partitioning
              fi
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

Set the kernel arguments.

  compute_kernel_args:
    type: OS::Heat::SoftwareConfig
    properties:
      config:
        str_replace:
          template: |
            #!/bin/bash
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              sed 's/^\(GRUB_CMDLINE_LINUX=".*\)"/\1 $KERNEL_ARGS isolcpus=$TUNED_CORES"/g' -i /etc/default/grub ;
              grub2-mkconfig -o /etc/grub2.cfg
              reboot
            fi
          params:
            $KERNEL_ARGS: {get_param: ComputeKernelArgs}
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}
            $TUNED_CORES: {get_param: HostIsolatedCoreList}

4.6.2. Modify post-install.yaml

Set the tuned configuration to enable CPU affinity.

  ExtraConfig:
    type: OS::Heat::SoftwareConfig
    properties:
      group: script
      config:
        str_replace:
          template: |
            #!/bin/bash

            set -x
            FORMAT=$COMPUTE_HOSTNAME_FORMAT
            if [[ -z $FORMAT ]] ; then
              FORMAT="compute" ;
            else
              # Assumption: only %index% and %stackname% are the variables in Host name format
              FORMAT=$(echo $FORMAT | sed  's/\%index\%//g' | sed 's/\%stackname\%//g') ;
            fi
            if [[ $(hostname) == *$FORMAT* ]] ; then
              tuned_service=/usr/lib/systemd/system/tuned.service
              grep -q "network.target" $tuned_service
              if [ "$?" -eq 0 ]; then
                sed -i '/After=.*/s/network.target//g' $tuned_service
              fi
              grep -q "Before=.*network.target" $tuned_service
              if [ ! "$?" -eq 0 ]; then
                grep -q "Before=.*" $tuned_service
                if [ "$?" -eq 0 ]; then
                  sed -i 's/^\(Before=.*\)/\1 network.target openvswitch.service/g' $tuned_service
                else
                  sed -i '/After/i Before=network.target openvswitch.service' $tuned_service
                fi
              fi
              systemctl daemon-reload
            fi
          params:
            $COMPUTE_HOSTNAME_FORMAT: {get_param: ComputeHostnameFormat}

4.6.3. Modify network-environment.yaml

Add the custom resources for OVS-DPDK under resource_registry.

resource_registry:
  # Specify the relative/absolute path to the config files you want to use for override the default.
  OS::TripleO::Compute::Net::SoftwareConfig: nic-configs/compute-ovs-dpdk.yaml
  OS::TripleO::Controller::Net::SoftwareConfig: nic-configs/controller.yaml
  OS::TripleO::NodeUserData: first-boot.yaml
  OS::TripleO::NodeExtraConfigPost: post-install.yaml

Under parameter_defaults, set the tunnel type and the tenant type to vxlan.
```
NeutronTunnelTypes: 'vxlan'
NeutronNetworkType: 'vxlan'
```
Under parameter_defaults, set the OVS-DPDK configuration parameters.
Note
NeutronDPDKCoreList and NeutronDPDKMemoryChannels are the required settings for this procedure. Attempting to deploy DPDK without appropriate values causes the deployment to fail or lead to unstable deployments.
1. Provide a list of cores that can be used as DPDK poll mode drivers (PMDs) in the format - [allowed_pattern: "[0-9,-]+"].
```
NeutronDpdkCoreList: "1,17,9,25"
```
  Note
  You must assign at least one CPU (with sibling thread) on each NUMA node with or without DPDK NICs present for DPDK PMD to avoid failures in creating guest instances.
To optimize OVS-DPDK performance, consider the following options:
- Select CPUs associated with the NUMA node of the DPDK interface. Use cat /sys/class/net/<interface>/device/numa_node to list the NUMA node associated with an interface and use lscpu to list the CPUs associated with that NUMA node.
- Group CPU siblings together (in case of hyper-threading). Use cat /sys/devices/system/cpu/<cpu>/topology/thread_siblings_list to find the sibling of a CPU.
- Reserve CPU 0 for the host process.
- Isolate CPUs assigned to PMD so that the host process does not use these CPUs.
- Use NovaVcpuPinset to exclude CPUs assigned to PMD from Compute scheduling.
  1. Provide the number of memory channels in the format - [allowed_pattern: "[0-9]+"].
    NeutronDpdkMemoryChannels: "4"
  2. Set the memory pre-allocated from the hugepage pool for each socket.
    NeutronDpdkSocketMemory: "2048,2048"
    This is a comma-separated string, in ascending order of the CPU socket. This example assumes a 2 NUMA node configuration and pre-allocates 2048 MB of huge pages to socket 0, and pre-allocates 2048 MB to socket 0. If you have a single NUMA node system, set this value to 1024,0.
  3. Set the DPDK driver type and the datapath type for OVS bridges.
    NeutronDpdkDriverType: "vfio-pci" NeutronDatapathType: "netdev"
Under parameter_defaults, set the vhost-user socket directory for OVS.
```
  NeutronVhostuserSocketDir: "/var/lib/vhost_sockets"
```
Under parameter_defaults, reserve the RAM for the host processes.
```
NovaReservedHostMemory: 2048
```
Under parameter_defaults, set a comma-separated list or range of physical CPU cores to reserve for virtual machine processes.
```
NovaVcpuPinSet: "2,3,4,5,6,7,18,19,20,21,22,23,10,11,12,13,14,15,26,27,28,29,30,31"
```
Under parameter_defaults, list the applicable filters.
Nova scheduler applies these filters in the order they are listed. List the most restrictive filters first to make the filtering process for the nodes more efficient.
```
NovaSchedulerDefaultFilters: "RamFilter,ComputeFilter,AvailabilityZoneFilter,ComputeCapabilitiesFilter,ImagePropertiesFilter,PciPassthroughFilter,NUMATopologyFilter"
```
Under parameter_defaults, define the ComputeKernelArgs parameters to be included in the default grub file at first boot.
```
ComputeKernelArgs: "default_hugepagesz=1GB hugepagesz=1G hugepages=32 iommu=pt intel_iommu=on"
```
Note
These huge pages are consumed by the virtual machines, and also by OVS-DPDK using the NeutronDpdkSocketMemory parameter as shown in this procedure. The number of huge pages available for the virtual machines is the boot parameter minus the NeutronDpdkSocketMemory.
You need to add hw:mem_page_size=1GB to the flavor you associate with the DPDK instance. If you do not do this, the instance does not get a DHCP allocation.
Under parameter_defaults, set a list or range of physical CPU cores to be isolated from the host.
The given argument is appended to the tuned cpu-partitioning profile.
```
HostIsolatedCoreList: "1,2,3,4,5,6,7,9,10,17,18,19,20,21,22,23,11,12,13,14,15,25,26,27,28,29,30,31"
```
Under parameter_defaults, set a list of logical cores used by OVS-DPDK processes for dpdk-lcore-mask. These cores must be mutually exclusive from the list of cores in NeutronDpdkCoreList and NovaVcpuPinSet.
```
HostCpusList: "0,16,8,24"
```

4.6.4. Modify controller.yaml

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageMgmtNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageMgmtIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: ExternalNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: ExternalIpSubnet}

Create the OVS bridge to the Compute node.

-
  type: ovs_bridge
  name: br-link
  use_dhcp: false
  members:
    -
      type: interface
      name: nic4
    -
      type: vlan
      vlan_id: {get_param: TenantNetworkVlanID}
      addresses:
        -
          ip_netmask: {get_param: TenantIpSubnet}

4.6.5. Modify compute.yaml

Create the compute-ovs-dpdk.yaml file from the default compute.yaml file and make the following changes:

Create the control plane Linux bond for an isolated network.

  -
    type: linux_bond
    name: bond_api
    bonding_options: "mode=active-backup"
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic2
        primary: true
      -
        type: interface
        name: nic3

Assign VLANs to this Linux bond.

  -
    type: vlan
    vlan_id: {get_param: InternalApiNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: InternalApiIpSubnet}
  -
    type: vlan
    vlan_id: {get_param: StorageNetworkVlanID}
    device: bond_api
    addresses:
      -
        ip_netmask: {get_param: StorageIpSubnet}

Set a bridge with a DPDK port to link to the controller.

  -
    type: ovs_user_bridge
    name: br-link
    use_dhcp: false
    ovs_extra:
      -
        str_replace:
          template: set port br-link tag=_VLAN_TAG_
          params:
            _VLAN_TAG_: {get_param: TenantNetworkVlanID}
    addresses:
      -
        ip_netmask: {get_param: TenantIpSubnet}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk0
        members:
          -
            type: interface
            name: nic4

Note

To include multiple DPDK devices, repeat the type code section for each DPDK device you want to add.

Note

4.6.6. Run the overcloud_deploy.sh Script

The following example defines the openstack overcloud deploy command for the OVS-DPDK environment within a Bash script:

#!/bin/bash

openstack overcloud deploy \
--templates \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml \
-e /home/stack/ospd-11-vxlan-dpdk-single-port-ctlplane-bonding/network-environment.yaml

/usr/share/openstack-tripleo-heat-templates/environments/neutron-ovs-dpdk.yaml is the location of the default neutron-ovs-dpdk.yaml file, which enables the OVS-DPDK parameters for the Compute role.
/home/stack/<relative-directory>/network-environment.yaml is the path for the network-environment.yaml file. Use this file to overwrite the default values from the neutron-ovs-dpdk.yaml file.

Note

Reboot the Compute nodes to enforce the tuned profile after the overcloud is deployed.

Note

This configuration of OVS-DPDK does not support security groups and live migrations.

4.7. Configure OVS-DPDK Composable Role

This section describes how to configure composable roles for OVS-DPDK and SR-IOV with VLAN tunnelling. The process to create and deploy a composable role includes:

Defining the new role in a local copy of the role_data.yaml file.
Creating the OpenStack flavor that uses this new role.
Modifying the network_environment.yaml file to include this new role.
Deploying the overcloud with this updated set of roles.

In this example, the ComputeOvsDpdk and ComputeSriov are composable role for compute nodes to enable DPDK or SR-IOV only on the nodes that have the SR-IOV NICs. The existing set of default roles provided by the Red Hat OpenStack Platform is stored in the /home/stack/roles_data.yaml file.

4.7.1. Modify roles_data.yaml to Create Composable Roles

Copy the roles_data.yaml file to your /home/stack/templates directory and add the new ComputeOvsDpdk and `ComputeSriov roles.

Define the OVS-DPDK composable role.

  - name: ComputeOvsDpdk
    CountDefault: 1
    HostnameFormatDefault: compute-ovs-dpdk-%index%
    disable_upgrade_deployment: True
    ServicesDefault:
      - OS::TripleO::Services::CACerts
      - OS::TripleO::Services::CephClient
      - OS::TripleO::Services::CephExternal
      - OS::TripleO::Services::Timezone
      - OS::TripleO::Services::Ntp
      - OS::TripleO::Services::Snmp
      - OS::TripleO::Services::Sshd
      - OS::TripleO::Services::NovaCompute
      - OS::TripleO::Services::NovaLibvirt
      - OS::TripleO::Services::Kernel
      - OS::TripleO::Services::ComputeNeutronCorePlugin
      - OS::TripleO::Services::ComputeNeutronOvsDpdkAgent
      - OS::TripleO::Services::ComputeCeilometerAgent
      - OS::TripleO::Services::ComputeNeutronL3Agent
      - OS::TripleO::Services::ComputeNeutronMetadataAgent
      - OS::TripleO::Services::TripleoPackages
      - OS::TripleO::Services::TripleoFirewall
      - OS::TripleO::Services::OpenDaylightOvs
      - OS::TripleO::Services::SensuClient
      - OS::TripleO::Services::FluentdClient
      - OS::TripleO::Services::AuditD
      - OS::TripleO::Services::Collectd

In this example, all the services of the new ComputeOvsDpdk role are same as that of a regular Compute role, with the exception of ComputeNeutronOvsAgent. Replace ComputeNeutronOvsAgent with ComputeNeutronOvsDpdkAgent to map to the OVS-DPDK service.

Define the SR-IOV composable role.

  - name: ComputeSriov
    CountDefault: 1
    HostnameFormatDefault: compute-sriov-%index%
    disable_upgrade_deployment: True
    ServicesDefault:
      - OS::TripleO::Services::CACerts
      - OS::TripleO::Services::CephClient
      - OS::TripleO::Services::CephExternal
      - OS::TripleO::Services::Timezone
      - OS::TripleO::Services::Ntp
      - OS::TripleO::Services::Snmp
      - OS::TripleO::Services::Sshd
      - OS::TripleO::Services::NovaCompute
      - OS::TripleO::Services::NovaLibvirt
      - OS::TripleO::Services::Kernel
      - OS::TripleO::Services::ComputeNeutronCorePlugin
      - OS::TripleO::Services::ComputeNeutronOvsAgent
      - OS::TripleO::Services::ComputeCeilometerAgent
      - OS::TripleO::Services::ComputeNeutronL3Agent
      - OS::TripleO::Services::ComputeNeutronMetadataAgent
      - OS::TripleO::Services::TripleoPackages
      - OS::TripleO::Services::TripleoFirewall
      - OS::TripleO::Services::NeutronSriovAgent
      - OS::TripleO::Services::OpenDaylightOvs
      - OS::TripleO::Services::SensuClient
      - OS::TripleO::Services::FluentdClient
      - OS::TripleO::Services::AuditD
      - OS::TripleO::Services::Collectd

4.7.2. Modify network-environment.yaml for the New Composable Roles

Add the resource mapping for the OVS-DPDK and SR-IOV services to the network-environment.yaml file along with the network configuration for this node as follows:

resource_registry:
  # Specify the relative/absolute path to the config files you want to use for override the default.
  OS::TripleO::ComputeSriov::Net::SoftwareConfig: nic-configs/compute-sriov.yaml
  OS::TripleO::Controller::Net::SoftwareConfig: nic-configs/controller.yaml
  OS::TripleO::ComputeOvsDpdk::Net::SoftwareConfig: nic-configs/compute-ovs-dpdk.yaml

  OS::TripleO::Services::ComputeNeutronOvsDpdkAgent: /usr/share/openstack-tripleo-heat-templates/puppet/services/neutron-ovs-dpdk-agent.yaml
  OS::TripleO::Services::NeutronSriovAgent: /usr/share/openstack-tripleo-heat-templates/puppet/services/neutron-sriov-agent.yaml

  OS::TripleO::NodeUserData: first-boot.yaml
  OS::TripleO::NodeExtraConfigPost: post-install.yaml

Configure the remainder of the network-environment.yaml file to override the default parameters from the neutron-ovs-dpdk-agent.yaml and neutron-sriov-agent.yaml files as needed for your OpenStack deployment.

4.7.3. Run the overcloud_deploy.sh Script

The following example defines the openstack overcloud deploy Bash script that uses composable roles:

# #!/bin/bash

openstack overcloud deploy \
--templates \
-r /home/stack/ospd-11-vlan-dpdk-sriov-single-port-composable-roles/roles-data.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml \
-e /home/stack/ospd-11-vlan-dpdk-sriov-single-port-composable-roles/network-environment.yaml

/home/stack/templates/roles_data.yaml is the location of the updated roles_data.yaml file, which defines the OVS-DPDK and the SR-IOV composable roles.

Note

Reboot the Compute nodes to enforce the tuned profile after the overcloud is deployed.

4.8. Set the MTU Value for OVS-DPDK Interfaces

Red Hat OpenStack Platform supports jumbo frames for OVS-DPDK. To set the MTU value for jumbo frames you must:

Set the global MTU value for networking in the network-environment.yaml file.
Set the physical DPDK port MTU value in the compute.yaml file. This value is also used by the vhost user interface.
Set the MTU value within any guest instances on the Compute node to ensure that you have a comparable MTU value from end to end in your configuration.

Note

VXLAN packets include an extra 50 bytes in the header. Calculate your MTU requirements based on these additional header bytes. For example, an MTU value of 9000 means the VXLAN tunnel MTU value is 8950 to account for these extra bytes.

Note

You do not need any special configuration for the physical NIC since the NIC is controlled by the DPDK PMD and has the same MTU value set by the compute.yaml file. You cannot set an MTU value larger than the maximum value supported by the physical NIC.

To set the MTU value for OVS-DPDK interfaces:

Set the NeutronGlobalPhysnetMtu parameter in the network-environment.yaml file.
```
parameter_defaults:
  # Global MTU configuration on Neutron
  NeutronGlobalPhysnetMtu: 2000
```
Note
Ensure that the NeutronDpdkSocketMemory value in the network-environment.yaml file is large enough to support jumbo frames. See Memory Parameters for details.

Set the MTU value on the bridge to the Compute node in the controller.yaml file.

  -
    type: ovs_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic4
        mtu: 2000
  -
    type: ovs_bridge
    name: br-link1
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: interface
        name: nic5
        mtu: 2000

Set the MTU value on the OVS-DPDK interfaces in the compute.yaml file.

  -
    type: ovs_user_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk0
        mtu: 2000
        ovs_extra:
        - set interface $DEVICE options:n_rxq=2
        - set interface $DEVICE mtu_request=$MTU
        members:
          -
            type: interface
            name: nic4
  -
    type: ovs_user_bridge
    name: br-link1
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk1
        mtu: 2000
        ovs_extra:
        - set interface $DEVICE options:n_rxq=2
        - set interface $DEVICE mtu_request=$MTU
        members:
          -
            type: interface
            name: nic5

4.9. Set Multiqueue for OVS-DPDK Interfaces

To set set same number of queues for interfaces in OVS-DPDK on the Compute node, modify the compute.yaml file as follows:

  -
    type: ovs_user_bridge
    name: br-link0
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk0
        mtu: 2000
        ovs_extra:
        - set interface $DEVICE options:n_rxq=2
        - set interface $DEVICE mtu_request=$MTU
        members:
          -
            type: interface
            name: nic4
  -
    type: ovs_user_bridge
    name: br-link1
    use_dhcp: false
    dns_servers: {get_param: DnsServers}
    members:
      -
        type: ovs_dpdk_port
        name: dpdk1
        mtu: 2000
        ovs_extra:
        - set interface $DEVICE options:n_rxq=2
        - set interface $DEVICE mtu_request=$MTU
        members:
          -
            type: interface
            name: nic5

4.10. Known Limitations

There are certain limitations when configuring OVS-DPDK with Red Hat OpenStack Platform 11 for the NFV use case:

Use Linux bonds for control plane networks. Ensure both PCI devices used in the bond are on the same NUMA node for optimum performance. Neutron Linux bridge configuration is not supported by Red Hat.
Huge pages are required for every instance running on the hosts with OVS-DPDK. If huge pages are not present in the guest, the interface appears but does not function.
There is a performance degradation of services that use tap devices, because these devices do not support DPDK. For example, services such as DVR, FWaaS, and LBaaS use tap devices.
- With OVS-DPDK, you can enable DVR with netdev datapath, but this has poor performance and is not suitable for a production environment. DVR uses kernel namespace and tap devices to perform the routing.
- To ensure the DVR routing performs well with OVS-DPDK, you need to use a controller such as ODL which implements routing as OpenFlow rules. With OVS-DPDK, OpenFlow routing removes the bottleneck introduced by the Linux kernel interfaces so that the full performance of datapath is maintained.
When using OVS-DPDK, all bridges on the same Compute node should be of type ovs_user_bridge. The director may accept the configuration, but Red Hat OpenStack Platform does not support mixing ovs_bridge and ovs_user_bridge on the same node.
Do not configure OVS-DPDK and SR-IOV on the same Compute node.

4.11. Create a Flavor and Deploy an Instance for OVS-DPDK

After you have completed configuring OVS-DPDK for your Red Hat OpenStack Platform deployment with NFV, you can create a flavor and deploy an instance with the following steps:

Create an aggregate group and add a host to it for OVS-DPDK.

 # openstack aggregate create --zone=dpdk dpdk
 # openstack aggregate add host dpdk compute-ovs-dpdk-0.localdomain

Create a flavor.
```
# openstack flavor create m1.medium_huge_4cpu --ram 4096 --disk 150 --vcpus 4
```
Here, m1.medium_huge_4cpu is the flavor name, 4096 is the memory size in MB, 150 is the disk size in GB (default 0G), and 4 is the number of vCPUs.
Set additional flavor properties.
```
# openstack flavor set --property hw:cpu_policy=dedicated --property hw:mem_page_size=large m1.medium_huge_4cpu
```
Here, m1.medium_huge_4cpu is the flavor name and the remaining parameters set the other properties for the flavor.

Create the network.

# openstack network create net1 --provider-physical-network tenant --provider-network-type vlan --provider-segment <VLAN-ID>

Deploy an instance.
```
# openstack server create --flavor m1.medium_huge_4cpu --availability-zone dpdk --image rhel_7.3 --nic net-id=net1
```
Where:
- m1.medium_huge_4cpu is the flavor name or ID.
- dpdk is the availability zone for the server.
- rhel_7.3 is the image (name or ID) used to create an instance.
- net1 is the NIC on the server.

You have now deployed an instance for the OVS-DPDK with NFV use case.

For using multi-queue with OVS-DPDK, there are a couple of additional steps that you need to include in the above procedure. Before you create a flavor, perform the following steps:

Set the image properties.
```
# openstack image set --property hw_vif_multiqueue_enabled=true <image-id>
```
Here, hw_vif_multiqueue_enabled=true is a property on this image to enable multiqueue, <image-id> is the name or ID of the image to modify.
Set additional flavor properties.
```
# openstack flavor set m1.vm_mq set hw:vif_multiqueue_enabled=true
```
Here, m1.vm_mq is the flavor ID or name, and the remaining options enable multiqueue for the flavor.

4.12. Optimizing Performance with Emulator Thread Pinning

To improve performance, you can pin the Qemu emulator thread to an alternate core.

Determine which cores are used as vCPUs for your instance:

# virsh dumpxml dpdk_vm | grep cpuset
    <vcpupin vcpu='0' cpuset='2'/>
    <vcpupin vcpu='1' cpuset='18'/>
    <vcpupin vcpu='2' cpuset='1'/>
    <vcpupin vcpu='3' cpuset='17'/>
    <emulatorpin cpuset='1-2,17-18'/>

Select the core you want to pin the emulator thread to. Ensure the selected core is from the NovaVcpuPinSet:
```
#virsh emulatorpin <vm-name> --cpulist 2
```
Note
The pCPU associated with the emulator pin thread consumes one vCPU (two threads if hyperthreading is enabled) from the NovaVcpuPinSet.

4.13. Troubleshooting the Configuration

This section describes the steps to troubleshoot the DPDK-OVS configuration.

Review the bridge configuration and confirm that the bridge was created with the datapath_type=netdev.

# ovs-vsctl list bridge br0
_uuid               : bdce0825-e263-4d15-b256-f01222df96f3
auto_attach         : []
controller          : []
datapath_id         : "00002608cebd154d"
datapath_type       : netdev
datapath_version    : "<built-in>"
external_ids        : {}
fail_mode           : []
flood_vlans         : []
flow_tables         : {}
ipfix               : []
mcast_snooping_enable: false
mirrors             : []
name                : "br0"
netflow             : []
other_config        : {}
ports               : [52725b91-de7f-41e7-bb49-3b7e50354138]
protocols           : []
rstp_enable         : false
rstp_status         : {}
sflow               : []
status              : {}
stp_enable          : false

Review the OVS service by confirming that the neutron-ovs-agent is configured to start automatically.

# systemctl status neutron-openvswitch-agent.service
neutron-openvswitch-agent.service - OpenStack Neutron Open vSwitch Agent
Loaded: loaded (/usr/lib/systemd/system/neutron-openvswitch-agent.service; enabled; vendor preset: disabled)
Active: active (running) since Mon 2015-11-23 14:49:31 AEST; 25min ago

If the service is having trouble starting, you can view any related messages.

# journalctl -t neutron-openvswitch-agent.service

Confirm that the PMD CPU mask of the ovs-dpdk are pinned to the CPUs. In case of HT, use sibling CPUs.

For example, take CPU4:

# cat /sys/devices/system/cpu/cpu4/topology/thread_siblings_list
4,20

So, using CPU 4 and 20:

# ovs-vsctl set Open_vSwitch . other_config:pmd-cpu-mask=0x100010

Display their status:

# tuna -t ovs-vswitchd -CP
thread  ctxt_switches pid SCHED_ rtpri affinity voluntary nonvoluntary       cmd
3161	OTHER 	0    	6	765023      	614	ovs-vswitchd
3219   OTHER 	0    	6     	1        	0   	handler24
3220   OTHER 	0    	6     	1        	0   	handler21
3221   OTHER 	0    	6     	1        	0   	handler22
3222   OTHER 	0    	6     	1        	0   	handler23
3223   OTHER 	0    	6     	1        	0   	handler25
3224   OTHER 	0    	6     	1        	0   	handler26
3225   OTHER 	0    	6     	1        	0   	handler27
3226   OTHER 	0    	6     	1        	0   	handler28
3227   OTHER 	0    	6     	2        	0   	handler31
3228   OTHER 	0    	6     	2        	4   	handler30
3229   OTHER 	0    	6     	2        	5   	handler32
3230   OTHER 	0    	6	953538      	431   revalidator29
3231   OTHER 	0    	6   1424258      	976   revalidator33
3232   OTHER 	0    	6   1424693      	836   revalidator34
3233   OTHER 	0    	6	951678      	503   revalidator36
3234   OTHER 	0    	6   1425128      	498   revalidator35
*3235   OTHER 	0    	4	151123       	51       	pmd37*
*3236   OTHER 	0   	20	298967       	48       	pmd38*
3164   OTHER 	0    	6 	47575        	0  dpdk_watchdog3
3165   OTHER 	0    	6	237634        	0   vhost_thread1
3166   OTHER 	0    	6  	3665        	0       	urcu2