1 .. This work is licensed under a Creative Commons Attribution 4.0 International
3 .. http://creativecommons.org/licenses/by/4.0
4 .. (c) OPNFV, Intel Corporation and others.
6 *************************************
7 Yardstick Test Case Description TC007
8 *************************************
10 .. _DPDKpktgen: https://github.com/Pktgen/Pktgen-DPDK/
11 .. _rfc2544: https://www.ietf.org/rfc/rfc2544.txt
13 +-----------------------------------------------------------------------------+
14 |Network Performance |
16 +--------------+--------------------------------------------------------------+
17 |test case id | OPNFV_YARDSTICK_TC007_Virtual Traffic Classifier Data Plane |
18 | | Throughput Benchmarking Test in Presence of Noisy |
21 +--------------+--------------------------------------------------------------+
22 |metric | Throughput |
24 +--------------+--------------------------------------------------------------+
25 |test purpose | To measure the throughput supported by the virtual Traffic |
26 | | Classifier according to the RFC2544 methodology for a |
27 | | user-defined set of vTC deployment configurations in the |
28 | | presence of noisy neighbours. |
30 +--------------+--------------------------------------------------------------+
31 |configuration | file: opnfv_yardstick_tc007.yaml |
33 | | packet_size: size of the packets to be used during the |
34 | | throughput calculation. |
35 | | Allowe values: [64, 128, 256, 512, 1024, 1280, 1518] |
37 | | vnic_type: type of VNIC to be used. |
38 | | Allowed values are: |
39 | | - normal: for default OvS port configuration |
40 | | - direct: for SR-IOV port configuration |
42 | | vtc_flavor: OpenStack flavor to be used for the vTC |
43 | | Default available values are: m1.small, m1.medium, |
44 | | and m1.large, but the user can create his/her own |
45 | | flavor and give it as input |
47 | | num_of_neighbours: Number of noisy neighbours (VMs) to be |
48 | | instantiated during the experiment. |
49 | | Allowed values: range (1, 10) |
51 | | amount_of_ram: RAM to be used by each neighbor. |
52 | | Allowed values: ['250M', '1G', '2G', '3G', '4G', '5G', |
53 | | '6G', '7G', '8G', '9G', '10G'] |
54 | | Deault value: 256M |
56 | | number_of_cores: Number of noisy neighbours (VMs) to be |
57 | | instantiated during the experiment. |
58 | | Allowed values: range (1, 10) |
59 | | Default value: 1 |
61 | | vlan_sender: vlan tag of the network on which the vTC will |
62 | | receive traffic (VLAN Network 1). |
63 | | Allowed values: range (1, 4096) |
65 | | vlan_receiver: vlan tag of the network on which the vTC |
66 | | will send traffic back to the packet generator |
67 | | (VLAN Network 2). |
68 | | Allowed values: range (1, 4096) |
70 | | default_net_name: neutron name of the defaul network that |
71 | | is used for access to the internet from the vTC |
74 | | default_subnet_name: subnet name for vNIC1 |
75 | | (information available through Neutron). |
77 | | vlan_net_1_name: Neutron Name for VLAN Network 1 |
78 | | (information available through Neutron). |
80 | | vlan_subnet_1_name: Subnet Neutron name for VLAN Network 1 |
81 | | (information available through Neutron). |
83 | | vlan_net_2_name: Neutron Name for VLAN Network 2 |
84 | | (information available through Neutron). |
86 | | vlan_subnet_2_name: Subnet Neutron name for VLAN Network 2 |
87 | | (information available through Neutron). |
89 +--------------+--------------------------------------------------------------+
90 |test tool | DPDK pktgen |
92 | | DPDK Pktgen is not part of a Linux distribution, |
93 | | hence it needs to be installed by the user. |
95 +--------------+--------------------------------------------------------------+
96 |references | DPDKpktgen_ |
102 +--------------+--------------------------------------------------------------+
103 |applicability | Test can be configured with different flavors, vNIC type |
104 | | and packet sizes. Default values exist as specified above. |
105 | | The vNIC type and flavor MUST be specified by the user. |
107 +--------------+--------------------------------------------------------------+
108 |pre-test | The vTC has been successfully instantiated and configured. |
109 | | The user has correctly assigned the values to the deployment |
110 | | configuration parameters. |
112 | | - Multicast traffic MUST be enabled on the network. |
113 | | The Data network switches need to be configured in |
114 | | order to manage multicast traffic. |
115 | | - In the case of SR-IOV vNICs use, SR-IOV compatible NICs |
116 | | must be used on the compute node. |
117 | | - Yarsdtick needs to be installed on a host connected to the |
118 | | data network and the host must have 2 DPDK-compatible |
119 | | NICs. Proper configuration of DPDK and DPDK pktgen is |
120 | | required before to run the test case. |
121 | | (For further instructions please refer to the ApexLake |
122 | | documentation). |
124 +--------------+--------------------------------------------------------------+
125 |test sequence | Description and expected results |
127 +--------------+--------------------------------------------------------------+
128 |step 1 | The noisy neighbours are deployed as required by the user. |
130 +--------------+--------------------------------------------------------------+
131 |step 2 | The vTC is deployed, according to the configuration required |
134 +--------------+--------------------------------------------------------------+
135 |step 3 | The vTC is correctly deployed and configured as necessary. |
136 | | The initialization script has been correctly executed and |
137 | | the vTC is ready to receive and process the traffic. |
139 +--------------+--------------------------------------------------------------+
140 |step 4 | Test case is executed with the parameters specified by the |
145 | | The traffic is sent to the vTC using the maximum available |
148 +--------------+--------------------------------------------------------------+
149 |step 5 | The vTC instance forwards all the packets back to the |
150 | | packet generator for 60 seconds, as specified by RFC 2544. |
152 | | Steps 4 and 5 are executed different times with different |
153 | | with different traffic rates, in order to find the maximum |
154 | | supported traffic rate, accoring to the current definition |
155 | | of throughput in RFC 2544. |
157 +--------------+--------------------------------------------------------------+
158 |test verdict | The result of the test is a number between 0 and 100 which |
159 | | represents the throughput in terms of percentage of the |
160 | | available pktgen NIC bandwidth. |
162 +--------------+--------------------------------------------------------------+