diff --git a/browbeat-config.yaml b/browbeat-config.yaml
index a0358b556..17502cf9b 100644
--- a/browbeat-config.yaml
+++ b/browbeat-config.yaml
@@ -415,6 +415,53 @@ workloads:
       - 8
     times: 10
     scenarios:
+      # This is a scenario developed specifically for an environment with a mix of
+      # non-NFV compute nodes, ComputeOvsDpdk/ComputeOvsDpdkSriov nodes,
+      # and compute nodes with OvS Hardware Offload/SR-IOV+OvS Hardware Offload.
+      # It creates three different availability zones for the three different types of compute nodes
+      # and boots VMs in them accordingly.
+      # The scenario boots <times> number of VMs in the availability zone which has
+      # the least number of compute nodes and automatically calculates the number of
+      # VMs to boot in the other AZs based on the proportional_scale option.
+      - name: nova-boot-hybrid-computes
+        enabled: true
+        image_name: centos7
+        vanilla_flavor_name: m1.small
+        vanilla_phys_net: provider-0
+        sriov_phys_net: provider-1
+        # option to enable booting DPDK instances in the tests
+        boot_dpdk_vms: False
+        dpdk_flavor_name: m1.small
+        dpdk_phys_net: provider-0
+        # Management network type for the VM can be either tenant and sriov.
+        # The port on the management network should be used for general access to
+        # to the VM and not for dataplane traffic.
+        dpdk_management_nw_type: tenant
+        dpdk_hosts_group: ComputeOvsDpdk
+        # option to enable booting instances with hardware offload configured in the tests
+        boot_hw_offload_vms: False
+        hw_offload_flavor_name: m1.small
+        hw_offload_phys_net: provider-2
+        # Management network type for the VM can be either tenant and sriov.
+        # The port on the management network should be used for general access to
+        # to the VM and not for dataplane traffic.
+        hw_offload_management_nw_type: tenant
+        hw_offload_hosts_group: ComputeSriovOffload
+        # path to tripleo inventory file
+        tripleo_inventory_file: ansible/hosts.yml
+        num_tenants: 2
+        num_networks_per_tenant: 6
+        # This option will scale VMs proportionally on
+        # different compute node types. For example,
+        # if there is 1 SR-IOV+DPDK compute node, 1 SR-IOV+HWOL
+        # compute node and 3 non-NFV compute nodes, then
+        # 3 non-NFV VMs will be booted in each iteration, while
+        # 1 NFV VM of each type will be booted in an iteration.
+        # This option should be enabled only when the environment
+        # is stable, as it could boot a lot of VMs concurrently.
+        proportional_scale: False
+        nova_api_version: 2.74
+        file: rally/rally-plugins/nova/nova_boot_hybrid_computes.yml
       - name: nova-boot-in-batches-with-delay
         enabled: true
         image_name: cirro5
diff --git a/rally/rally-plugins/nova/nfv_context.py b/rally/rally-plugins/nova/nfv_context.py
new file mode 100644
index 000000000..4034128bb
--- /dev/null
+++ b/rally/rally-plugins/nova/nfv_context.py
@@ -0,0 +1,308 @@
+#   Licensed under the Apache License, Version 2.0 (the "License");
+#   you may not use this file except in compliance with the License.
+#   You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+#   Unless required by applicable law or agreed to in writing, software
+#   distributed under the License is distributed on an "AS IS" BASIS,
+#   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#   See the License for the specific language governing permissions and
+#   limitations under the License.
+
+from rally.task import context
+from rally.common import logging
+from rally.common import utils
+from rally import consts
+from rally_openstack.common import osclients
+from rally_openstack.common.wrappers import network as network_wrapper
+
+import yaml
+
+LOG = logging.getLogger(__name__)
+
+
+@context.configure(name="create_nfv_azs_and_networks", order=1100)
+class CreateNFVAZsandNetworksContext(context.Context):
+    """This plugin creates availability zones with host aggregates and networks
+    for non-NFV, SR-IOV with DPDK and SR-IOV with hardware offload.
+    """
+
+    CONFIG_SCHEMA = {
+        "type": "object",
+        "$schema": consts.JSON_SCHEMA,
+        "additionalProperties": True,
+        "properties": {
+            "boot_dpdk_vms": {
+                "type": "string",
+                "default": "False"
+            },
+            "boot_hw_offload_vms": {
+                "type": "string",
+                "default": "False"
+            },
+            "dpdk_hosts_group": {
+                "type": "string",
+                "default": "ComputeOvsDpdkSriov"
+            },
+            "hw_offload_hosts_group": {
+                "type": "string",
+                "default": "ComputeSriovOffload"
+            },
+            "tripleo_inventory_file": {
+                "type": "string",
+                "default": "/home/stack/browbeat/ansible/hosts.yml"
+            }
+        }
+    }
+
+    def _create_subnet(self, tenant_id, network_id, network_number):
+        """Create subnet for network
+
+        :param tenant_id: ID of tenant
+        :param network_id: ID of network
+        :param network_number: int, number for CIDR of subnet
+        :returns: subnet object
+        """
+        subnet_args = {
+            "subnet": {
+                "tenant_id": tenant_id,
+                "network_id": network_id,
+                "name": self.net_wrapper.owner.generate_random_name(),
+                "ip_version": 4,
+                "cidr": "172.{}.0.0/16".format(network_number),
+                "enable_dhcp": True,
+                "gateway_ip": "172.{}.0.1".format(network_number),
+                "allocation_pools": [{"start": "172.{}.0.2".format(network_number),
+                                      "end": "172.{}.254.254".format(network_number)}]
+            }
+        }
+        return self.net_wrapper.client.create_subnet(subnet_args)["subnet"]
+
+    def _create_networks_and_subnets(self, tenant_id):
+        self.net_wrapper = network_wrapper.wrap(
+            osclients.Clients(self.context["admin"]["credential"]),
+            self,
+            config=self.config,
+        )
+
+        net_kwargs = {
+            "network_create_args": {
+                "provider:network_type": "vlan",
+                "provider:physical_network": self.config["provider_phys_nets"]["vanilla"],
+                "provider:segmentation_id": self.network_vlan_number,
+            }
+        }
+        vanilla_network = self.net_wrapper.create_network(tenant_id, **net_kwargs)
+
+        self.context["vanilla_networks"][tenant_id] = vanilla_network
+        self.context["vanilla_subnets"][tenant_id] = self._create_subnet(
+            tenant_id, vanilla_network["id"],
+            self.network_vlan_number-950)
+        self.network_vlan_number += 1
+
+        if self.context["boot_dpdk_vms"]:
+            net_kwargs = {
+                "network_create_args": {
+                    "provider:network_type": "vlan",
+                    "provider:physical_network": self.config["provider_phys_nets"]["dpdk"],
+                    "provider:segmentation_id": self.network_vlan_number,
+                }
+            }
+
+            dpdk_network = self.net_wrapper.create_network(tenant_id, **net_kwargs)
+            self.context["nfv_networks"].setdefault("dpdk", {})
+            self.context["nfv_networks"]["dpdk"][tenant_id] = dpdk_network
+            self.context["nfv_subnets"].setdefault("dpdk", {})
+            self.context["nfv_subnets"]["dpdk"][tenant_id] = self._create_subnet(
+                tenant_id, dpdk_network["id"],
+                self.network_vlan_number-950)
+            self.network_vlan_number += 1
+
+            net_kwargs = {
+                "network_create_args": {
+                    "provider:network_type": "vlan",
+                    "provider:physical_network": self.config["provider_phys_nets"]["sriov"],
+                    "provider:segmentation_id": self.network_vlan_number,
+                }
+            }
+            sriov_network = self.net_wrapper.create_network(tenant_id, **net_kwargs)
+
+            self.context["nfv_networks"].setdefault("sriov", {})
+            self.context["nfv_networks"]["sriov"][tenant_id] = sriov_network
+            self.context["nfv_subnets"].setdefault("sriov", {})
+            self.context["nfv_subnets"]["sriov"][tenant_id] = self._create_subnet(
+                tenant_id, sriov_network["id"],
+                self.network_vlan_number-950)
+            self.network_vlan_number += 1
+
+        if self.context["boot_hw_offload_vms"]:
+            net_kwargs = {
+                "network_create_args": {
+                    "provider:network_type": "vlan",
+                    "provider:physical_network": self.config["provider_phys_nets"]["hw_offload"],
+                    "provider:segmentation_id": self.network_vlan_number,
+                }
+            }
+            hw_offload_network = self.net_wrapper.create_network(tenant_id, **net_kwargs)
+
+            self.context["nfv_networks"].setdefault("hw_offload", {})
+            self.context["nfv_networks"]["hw_offload"][tenant_id] = hw_offload_network
+            self.context["nfv_subnets"].setdefault("hw_offload", {})
+            self.context["nfv_subnets"]["hw_offload"][tenant_id] = self._create_subnet(
+                tenant_id, hw_offload_network["id"],
+                self.network_vlan_number-950)
+            self.network_vlan_number += 1
+
+    def setup(self):
+        """This method is called before the task starts."""
+        self.net_wrapper = network_wrapper.wrap(
+            osclients.Clients(self.context["admin"]["credential"]),
+            self,
+            config=self.config,
+        )
+        self.nova_wrapper = osclients.Nova(self.context["admin"]["credential"]).create_client()
+        self.context["nfv_networks"] = {}
+        self.context["nfv_subnets"] = {}
+        self.context["vanilla_networks"] = {}
+        self.context["vanilla_subnets"] = {}
+        # This has been made a string value so that upper case/lower case
+        # variations can be considered.
+        self.context["boot_dpdk_vms"] = self.config.get("boot_dpdk_vms", "False") in [
+            "True", "true"]
+        self.context["boot_hw_offload_vms"] = self.config.get("boot_hw_offload_vms",
+                                                              "False") in [
+                                                                  "True", "true"]
+        self.dpdk_hosts_group = self.config.get("dpdk_hosts_group",
+                                                "ComputeOvsDpdk")
+        self.hw_offload_hosts_group = self.config.get("hw_offload_hosts_group",
+                                                      "ComputeSriovOffload")
+        self.network_vlan_number = 1001
+
+        tripleo_inventory_file_path = self.config.get("tripleo_inventory_file",
+                                                      "/home/stack/browbeat/ansible/hosts.yml")
+        with open(tripleo_inventory_file_path, "r") as tripleo_inventory_file:
+            self.tripleo_inventory = yaml.safe_load(tripleo_inventory_file)
+
+        dpdk_and_hw_offload_hosts = []
+
+        if self.context["boot_dpdk_vms"]:
+            self.dpdk_aggregate = self.nova_wrapper.aggregates.create(
+                "dpdk_aggregate", "az_dpdk")
+            dpdk_hosts = self.tripleo_inventory[self.dpdk_hosts_group]["hosts"]
+            self.context["num_dpdk_compute_hosts"] = len(dpdk_hosts)
+            for host_details in dpdk_hosts.values():
+                self.nova_wrapper.aggregates.add_host(self.dpdk_aggregate.id,
+                                                      host_details["canonical_hostname"])
+                dpdk_and_hw_offload_hosts.append(host_details["canonical_hostname"])
+
+        if self.context["boot_hw_offload_vms"]:
+            self.hw_offload_aggregate = self.nova_wrapper.aggregates.create(
+                "hw_offload_aggregate", "az_hw_offload")
+            hw_offload_hosts = self.tripleo_inventory[
+                self.hw_offload_hosts_group]["hosts"]
+            self.context["num_hw_offload_compute_hosts"] = len(hw_offload_hosts)
+            for host_details in hw_offload_hosts.values():
+                self.nova_wrapper.aggregates.add_host(self.hw_offload_aggregate.id,
+                                                      host_details["canonical_hostname"])
+                dpdk_and_hw_offload_hosts.append(host_details["canonical_hostname"])
+
+        self.vanilla_compute_aggregate = self.nova_wrapper.aggregates.create(
+            "vanilla_compute_aggregate", "az_vanilla_compute")
+
+        self.vanilla_compute_hosts = set()
+        for hostsgroup in self.tripleo_inventory:
+            if "hosts" in self.tripleo_inventory[hostsgroup] and "Compute" in hostsgroup:
+                for host_details in self.tripleo_inventory[hostsgroup]["hosts"].values():
+                    if ("canonical_hostname" in host_details and
+                       host_details["canonical_hostname"] not in dpdk_and_hw_offload_hosts):
+                        self.nova_wrapper.aggregates.add_host(
+                            self.vanilla_compute_aggregate.id,
+                            host_details["canonical_hostname"])
+                        self.vanilla_compute_hosts.add(host_details["canonical_hostname"])
+        self.context["num_vanilla_compute_hosts"] = len(self.vanilla_compute_hosts)
+
+        for _, tenant_id in utils.iterate_per_tenants(
+            self.context.get("users", [])
+        ):
+            self._create_networks_and_subnets(tenant_id)
+
+    def cleanup(self):
+        """This method is called after the task finishes."""
+        for hostname in self.vanilla_compute_hosts:
+            self.nova_wrapper.aggregates.remove_host(self.vanilla_compute_aggregate.id,
+                                                     hostname)
+        self.nova_wrapper.aggregates.delete(self.vanilla_compute_aggregate.id)
+
+        if self.context["boot_dpdk_vms"]:
+            dpdk_hosts = self.tripleo_inventory[self.dpdk_hosts_group]["hosts"]
+            for host_details in dpdk_hosts.values():
+                self.nova_wrapper.aggregates.remove_host(self.dpdk_aggregate.id,
+                                                         host_details["canonical_hostname"])
+            self.nova_wrapper.aggregates.delete(self.dpdk_aggregate.id)
+
+        if self.context["boot_hw_offload_vms"]:
+            hw_offload_hosts = self.tripleo_inventory[
+                self.hw_offload_hosts_group]["hosts"]
+            for host_details in hw_offload_hosts.values():
+                self.nova_wrapper.aggregates.remove_host(self.hw_offload_aggregate.id,
+                                                         host_details["canonical_hostname"])
+            self.nova_wrapper.aggregates.delete(self.hw_offload_aggregate.id)
+
+        for subnet in self.context["vanilla_subnets"].values():
+            try:
+                subnet_id = subnet["id"]
+                self.net_wrapper._delete_subnet(subnet_id)
+                LOG.debug(
+                    "Subnet with id '%s' deleted from context"
+                    % subnet_id
+                )
+            except Exception as e:
+                msg = "Can't delete subnet {} from context: {}".format(
+                    subnet_id, e
+                )
+                LOG.warning(msg)
+
+        for network in self.context["vanilla_networks"].values():
+            try:
+                network_id = network["id"]
+                self.net_wrapper.delete_network(network)
+                LOG.debug(
+                    "Network with id '%s' deleted from context"
+                    % network_id
+                )
+            except Exception as e:
+                msg = "Can't delete network {} from context: {}".format(
+                    network_id, e
+                )
+                LOG.warning(msg)
+
+        for subnets in self.context["nfv_subnets"].values():
+            for subnet in subnets.values():
+                try:
+                    nfv_subnet_id = subnet["id"]
+                    self.net_wrapper._delete_subnet(nfv_subnet_id)
+                    LOG.debug(
+                        "Subnet with id '%s' deleted from context"
+                        % nfv_subnet_id
+                    )
+                except Exception as e:
+                    msg = "Can't delete subnet {} from context: {}".format(
+                        nfv_subnet_id, e
+                    )
+                    LOG.warning(msg)
+
+        for networks in self.context["nfv_networks"].values():
+            for network in networks.values():
+                try:
+                    nfv_network_id = network["id"]
+                    self.net_wrapper.delete_network(network)
+                    LOG.debug(
+                        "Network with id '%s' deleted from context"
+                        % nfv_network_id
+                    )
+                except Exception as e:
+                    msg = "Can't delete network {} from context: {}".format(
+                        nfv_network_id, e
+                    )
+                    LOG.warning(msg)
diff --git a/rally/rally-plugins/nova/nova.py b/rally/rally-plugins/nova/nova.py
index 22293fd56..c030dedcf 100644
--- a/rally/rally-plugins/nova/nova.py
+++ b/rally/rally-plugins/nova/nova.py
@@ -15,11 +15,15 @@ import time
 from rally_openstack.common import consts
 from rally_openstack.task.scenarios.cinder import utils as cinder_utils
 from rally_openstack.task.scenarios.nova import utils as nova_utils
+from rally_openstack.task.scenarios.neutron import utils as neutron_utils
 from rally_openstack.task.scenarios.vm import utils as vm_utils
 from rally.task import scenario
 from rally.task import types
 from rally.task import validation
 
+from rally_openstack.common.services.network import neutron
+from rally_openstack.common import osclients
+
 LOG = logging.getLogger(__name__)
 
 @types.convert(image={"type": "glance_image"}, flavor={"type": "nova_flavor"})
@@ -33,6 +37,166 @@ class NovaBootPersist(nova_utils.NovaScenario):
     def run(self, image, flavor, **kwargs):
         self._boot_server(image, flavor)
 
+@types.convert(image={"type": "glance_image"}, vanilla_flavor={"type": "nova_flavor"},
+               dpdk_flavor={"type": "nova_flavor"}, hw_offload_flavor={"type": "nova_flavor"})
+@validation.add("image_valid_on_flavor", flavor_param="vanilla_flavor", image_param="image")
+@validation.add("image_valid_on_flavor", flavor_param="dpdk_flavor", image_param="image")
+@validation.add("image_valid_on_flavor", flavor_param="hw_offload_flavor", image_param="image")
+@validation.add("required_contexts", contexts=("create_nfv_azs_and_networks"))
+@validation.add("required_services", services=[consts.Service.NEUTRON, consts.Service.NOVA])
+@validation.add("required_platform", platform="openstack", users=True)
+@scenario.configure(context={"cleanup@openstack": ["neutron", "nova"]},
+                    name="BrowbeatNova.nova_boot_hybrid_computes", platform="openstack")
+class NovaBootHybridComputes(nova_utils.NovaScenario, neutron_utils.NeutronScenario):
+
+    def run(self, image, vanilla_flavor, dpdk_flavor,
+            hw_offload_flavor, num_networks_per_tenant,
+            dpdk_management_nw_type, hw_offload_management_nw_type,
+            proportional_scale, **kwargs):
+        """Create VMs on non-NFV compute nodes, SR-IOV+DPDK compute nodes,
+           and SR-IOV+Hardware Offload compute nodes.
+        :param image: image ID or instance for server creation
+        :param vanilla_flavor: flavor ID or instance for server creation
+        :param dpdk_flavor: flavor ID or instance for SR-IOV+DPDK VM creation
+        :param hw_offload_flavor: flavor ID or instance for SR-IOV+Hardware Offload VM creation
+        :param num_networks_per_tenant: int, number of tunnel networks per tenant
+        :param dpdk_management_nw_type: str, management network for DPDK VMs
+        :param hw_offload_management_nw_type: str, management network for HW Offload VMs
+        :param proportional_scale: str, option to scale VMs proportionately
+        """
+        tenant_id = self.context["tenant"]["id"]
+        minimum_num_compute_hosts = min(self.context["num_vanilla_compute_hosts"],
+                                        self.context.get("num_dpdk_compute_hosts",
+                                                         self.context[
+                                                             "num_vanilla_compute_hosts"]),
+                                        self.context.get("num_hw_offload_compute_hosts",
+                                                         self.context[
+                                                             "num_vanilla_compute_hosts"]))
+
+        vanilla_provider_network = self.context["vanilla_networks"][tenant_id]
+        tenant_network_id = self.context["tenant"]["networks"][((self.context["iteration"]-1)
+                                                               % num_networks_per_tenant)]["id"]
+        LOG.info("ITER {} using tenant network {}".format(self.context["iteration"],
+                                                          tenant_network_id))
+        kwargs["nics"] = [{'net-id': vanilla_provider_network["id"]}]
+        kwargs["availability-zone"] = "az_vanilla_compute"
+
+        proportional_scale = proportional_scale in ["True", "true"]
+        if proportional_scale:
+            num_vanilla_vms_to_boot = self.context[
+                "num_vanilla_compute_hosts"] // minimum_num_compute_hosts
+            if self.context["boot_dpdk_vms"]:
+                num_dpdk_vms_to_boot = self.context[
+                    "num_dpdk_compute_hosts"] // minimum_num_compute_hosts
+            if self.context["boot_hw_offload_vms"]:
+                num_sriov_hw_offload_vms_to_boot = self.context[
+                    "num_hw_offload_compute_hosts"] // minimum_num_compute_hosts
+        else:
+            num_vanilla_vms_to_boot = 1
+            if self.context["boot_dpdk_vms"]:
+                num_dpdk_vms_to_boot = 1
+            if self.context["boot_hw_offload_vms"]:
+                num_sriov_hw_offload_vms_to_boot = 1
+
+        for _ in range(num_vanilla_vms_to_boot):
+            vanilla_server = self._boot_server(image, vanilla_flavor, **kwargs)
+            self._attach_interface(vanilla_server, net_id=tenant_network_id)
+            LOG.info("ITER {} Booted vanilla server : {}".format(self.context["iteration"],
+                                                                 vanilla_server.id))
+            # VMs booting simultaneously across iterations adds a lot of load, so delay booting VMs
+            # for 5 seconds.
+            time.sleep(5)
+
+        if self.context["boot_dpdk_vms"]:
+            LOG.info("ITER {} DPDK instances enabled.".format(self.context["iteration"]))
+
+            dpdk_server_kwargs = {}
+            dpdk_networks = self.context["nfv_networks"]["dpdk"]
+
+            if dpdk_management_nw_type == "sriov":
+                sriov_port_kwargs = {}
+                sriov_networks = self.context["nfv_networks"]["sriov"]
+                sriov_network = sriov_networks[tenant_id]
+                sriov_port_kwargs["binding:vnic_type"] = "direct"
+
+            for _ in range(num_dpdk_vms_to_boot):
+                if dpdk_management_nw_type == "sriov":
+                    sriov_port = self._create_port({"network": sriov_network}, sriov_port_kwargs)
+                    dpdk_server_kwargs["nics"] = [{'port-id': sriov_port["port"]["id"]},
+                                                  {'net-id': dpdk_networks[tenant_id]["id"]}]
+                elif dpdk_management_nw_type == "tenant":
+                    dpdk_server_kwargs["nics"] = [{'net-id': tenant_network_id},
+                                                  {'net-id': dpdk_networks[tenant_id]["id"]}]
+                else:
+                    raise Exception("{} is not a valid management network type. {}".format(
+                                    dpdk_management_nw_type, "Please choose sriov or tenant."))
+                dpdk_server_kwargs["availability-zone"] = "az_dpdk"
+                dpdk_server = self._boot_server(image, dpdk_flavor,
+                                                **dpdk_server_kwargs)
+                LOG.info("ITER {} Booted DPDK server : {}".format(self.context["iteration"],
+                                                                  dpdk_server.id))
+                # VMs booting simultaneously across iterations adds a lot of load,
+                # so delay booting VMs for 5 seconds.
+                time.sleep(5)
+
+        if self.context["boot_hw_offload_vms"]:
+            LOG.info("ITER {} Hardware Offload Instances enabled.".format(
+                     self.context["iteration"]))
+
+            hw_offload_server_kwargs = {}
+            hw_offload_network = self.context["nfv_networks"]["hw_offload"][tenant_id]
+            hw_offload_subnet = self.context["nfv_subnets"]["hw_offload"][tenant_id]
+
+            if hw_offload_management_nw_type == "sriov":
+                sriov_networks = self.context["nfv_networks"]["sriov"]
+                sriov_network = sriov_networks[tenant_id]
+                sriov_port_kwargs = {}
+                sriov_port_kwargs["binding:vnic_type"] = "direct"
+
+            admin_clients = osclients.Clients(self.context["admin"]["credential"])
+            self.admin_neutron = neutron.NeutronService(
+                clients=admin_clients,
+                name_generator=self.generate_random_name,
+                atomic_inst=self.atomic_actions()
+            )
+
+            hw_offload_port_kwargs = {}
+            hw_offload_port_kwargs["binding:vnic_type"] = "direct"
+            hw_offload_port_kwargs["fixed_ips"] = [{"subnet_id": hw_offload_subnet["id"]}]
+
+            for _ in range(num_sriov_hw_offload_vms_to_boot):
+                if hw_offload_management_nw_type == "sriov":
+                    sriov_port = self._create_port({"network": sriov_network}, sriov_port_kwargs)
+                hw_offload_port = self._create_port({"network": hw_offload_network},
+                                                    hw_offload_port_kwargs)
+
+                hw_offload_port_kwargs["binding:profile"] = {"capabilities": ["switchdev"]}
+                hw_offload_port = {"port": self.admin_neutron.update_port(
+                                   port_id=hw_offload_port["port"]["id"],
+                                   **hw_offload_port_kwargs)}
+
+                if hw_offload_management_nw_type == "sriov":
+                    hw_offload_server_kwargs["nics"] = [{'port-id': sriov_port["port"]["id"]},
+                                                        {'port-id':
+                                                            hw_offload_port["port"]["id"]}]
+                elif hw_offload_management_nw_type == "tenant":
+                    hw_offload_server_kwargs["nics"] = [{'net-id': tenant_network_id},
+                                                        {'port-id':
+                                                            hw_offload_port["port"]["id"]}]
+                else:
+                    raise Exception("{} is not a valid management network type. {}".format(
+                                    hw_offload_management_nw_type,
+                                    "Please choose sriov or tenant."))
+
+                hw_offload_server_kwargs["availability-zone"] = "az_hw_offload"
+                hw_offload_server = self._boot_server(image, hw_offload_flavor,
+                                                      **hw_offload_server_kwargs)
+                LOG.info("ITER {} Booted Hardware Offload server : {}".format(
+                         self.context["iteration"], hw_offload_server.id))
+
+                # VMs booting simultaneously across iterations adds a lot of load,
+                # so delay booting VMs for 5 seconds.
+                time.sleep(5)
 
 @types.convert(image={"type": "glance_image"}, flavor={"type": "nova_flavor"})
 @validation.add("image_valid_on_flavor", flavor_param="flavor", image_param="image")
diff --git a/rally/rally-plugins/nova/nova_boot_hybrid_computes.yml b/rally/rally-plugins/nova/nova_boot_hybrid_computes.yml
new file mode 100644
index 000000000..b913d03c6
--- /dev/null
+++ b/rally/rally-plugins/nova/nova_boot_hybrid_computes.yml
@@ -0,0 +1,66 @@
+{% set image_name = image_name or 'centos7' %}
+{% set vanilla_flavor_name = vanilla_flavor_name or 'm1.small' %}
+{% set dpdk_flavor_name = dpdk_flavor_name or 'm1.small' %}
+{% set hw_offload_flavor_name = hw_offload_flavor_name or 'm1.small' %}
+{% set proportional_scale = proportional_scale or 'False' %}
+{% set nova_api_version = nova_api_version or 2.74 %}
+{% set num_tenants = num_tenants or 1 %}
+{% set num_networks_per_tenant = num_networks_per_tenant or 1 %}
+{% set sla_max_avg_duration = sla_max_avg_duration or 60 %}
+{% set sla_max_failure = sla_max_failure or 0 %}
+{% set sla_max_seconds = sla_max_seconds or 60 %}
+---
+BrowbeatNova.nova_boot_hybrid_computes:
+  -
+    args:
+      image:
+        name: '{{ image_name }}'
+      vanilla_flavor:
+        name: '{{ vanilla_flavor_name }}'
+      dpdk_flavor:
+        name: '{{ dpdk_flavor_name }}'
+      hw_offload_flavor:
+        name: '{{ hw_offload_flavor_name }}'
+      dpdk_management_nw_type: '{{ dpdk_management_nw_type }}'
+      hw_offload_management_nw_type: '{{ hw_offload_management_nw_type }}'
+      proportional_scale: '{{ proportional_scale }}'
+      num_networks_per_tenant: {{ num_networks_per_tenant }}
+    runner:
+      concurrency: {{concurrency}}
+      times: {{times}}
+      type: 'constant'
+    context:
+      create_nfv_azs_and_networks:
+        provider_phys_nets:
+          dpdk: '{{ dpdk_phys_net }}'
+          sriov: '{{ sriov_phys_net }}'
+          hw_offload: '{{ hw_offload_phys_net }}'
+          vanilla: '{{ vanilla_phys_net }}'
+        boot_dpdk_vms: '{{ boot_dpdk_vms }}'
+        boot_hw_offload_vms: '{{ boot_hw_offload_vms }}'
+        dpdk_hosts_group: '{{ dpdk_hosts_group }}'
+        hw_offload_hosts_group: '{{ hw_offload_hosts_group }}'
+        tripleo_inventory_file: '{{ tripleo_inventory_file }}'
+      api_versions:
+        nova:
+          version: {{ nova_api_version }}
+      users:
+        tenants: {{ num_tenants }}
+        users_per_tenant: 8
+      network:
+        networks_per_tenant: {{num_networks_per_tenant}}
+      quotas:
+        neutron:
+          network: -1
+          port: -1
+          subnet: -1
+          router: -1
+        nova:
+          instances: -1
+          cores: -1
+          ram: -1
+    sla:
+      max_avg_duration: {{sla_max_avg_duration}}
+      max_seconds_per_iteration: {{sla_max_seconds}}
+      failure_rate:
+        max: {{sla_max_failure}}