NFVBENCH-134 Support multiple idle interfaces per test VM

Enhance test VM to support auto-selection of PCI addresses based on MAC address
Enhance syslog reporting in test VM
Increase test VM version to 0.7
Add new properties to cfg.defaults.yaml to support idle interfaces
Enhance cleanup to support idle networks and ports

Change-Id: Ia89a2a1bee9eb5fb56f7adde074bdcec09cb0a3f
Signed-off-by: ahothan <ahothan@cisco.com>

1 files changed, 32 insertions, 0 deletions

diff --git a/nfvbench/cfg.default.yaml b/nfvbench/cfg.default.yaml
index 3152bb9..eb5fa11 100755
--- a/nfvbench/cfg.default.yaml
+++ b/nfvbench/cfg.default.yaml
@@ -385,6 +385,38 @@ internal_networks:
         segmentation_id:
         physical_network:
 
+# IDLE INTERFACES: PVP, PVVP and non shared net only.
+# By default each test VM will have 2 virtual interfaces for looping traffic.
+# If service_chain_shared_net is false, additional virtual interfaces can be
+# added at VM creation time, these interfaces will not carry any traffic and
+# can be used to test the impact of idle interfaces in the overall performance.
+# All these idle interfaces will use normal ports (not direct).
+# Number of idle interfaces per VM (none by default)
+idle_interfaces_per_vm: 0
+
+# A new network is created for each idle interface.
+# If service_chain_shared_net is true, the options below will be ignored
+# and no idle interfaces will be added.
+idle_networks:
+    # Prefix for all idle networks
+    name: 'nfvbench-idle-net'
+    # Prefix for all idle subnetworks 
+    subnet: 'nfvbench-idle-subnet'
+    # CIDR to use for all idle networks (value should not matter)
+    cidr: '192.169.1.0/24'
+    # Type of network associated to the idle virtual interfaces (vlan or vxlan)
+    network_type: 'vlan'
+    # segmentation ID to use for the network attached to the idle virtual interfaces
+    # vlan: leave empty to let neutron pick the segmentation ID
+    # vxlan: must specify the VNI value to be used (cannot be empty)
+    # Note that NFVbench will use as many consecutive segmentation IDs as needed.
+    # For example, for 4 PVP chains and 8 idle
+    # interfaces per VM, NFVbench will use 32 consecutive values of segmentation ID
+    # starting from the value provided.
+    segmentation_id:
+    # physnet name to use for all idle interfaces
+    physical_network:
+
 # In the scenario of PVVP + SRIOV, there is choice of how the traffic will be
 # handled in the middle network. The default (false) will use vswitch, while
 # SRIOV can be used by toggling below setting.