docs/testing/user/userguide/opnfv_yardstick_tc050.rst

   1 .. This work is licensed under a Creative Commons Attribution 4.0 International
   2 .. License.
   3 .. http://creativecommons.org/licenses/by/4.0
   4 .. (c) OPNFV, Yin Kanglin and others.
   5 .. 14_ykl@tongji.edu.cn
   6
   7 *************************************
   8 Yardstick Test Case Description TC050
   9 *************************************
  10
  11 +-----------------------------------------------------------------------------+
  12 |OpenStack Controller Node Network High Availability                          |
  13 |                                                                             |
  14 +--------------+--------------------------------------------------------------+
  15 |test case id  | OPNFV_YARDSTICK_TC050: OpenStack Controller Node Network     |
  16 |              | High Availability                                            |
  17 +--------------+--------------------------------------------------------------+
  18 |test purpose  | This test case will verify the high availability of control  |
  19 |              | node. When one of the controller failed to connect the       |
  20 |              | network, which breaks down the Openstack services on this    |
  21 |              | node. These Openstack service should able to be accessed by  |
  22 |              | other controller nodes, and the services on failed           |
  23 |              | controller node should be isolated.                          |
  24 +--------------+--------------------------------------------------------------+
  25 |test method   | This test case turns off the network interfaces of a         |
  26 |              | specified control node, then checks whether all services     |
  27 |              | provided by the control node are OK with some monitor tools. |
  28 +--------------+--------------------------------------------------------------+
  29 |attackers     | In this test case, an attacker called "close-interface" is   |
  30 |              | needed. This attacker includes three parameters:             |
  31 |              | 1) fault_type: which is used for finding the attacker's      |
  32 |              | scripts. It should be always set to "close-interface" in     |
  33 |              | this test case.                                              |
  34 |              | 2) host: which is the name of a control node being attacked. |
  35 |              | 3) interface: the network interface to be turned off.        |
  36 |              |                                                              |
  37 |              | There are four instance of the "close-interface" monitor:    |
  38 |              | attacker1(for public netork):                                |
  39 |              | -fault_type: "close-interface"                               |
  40 |              | -host: node1                                                 |
  41 |              | -interface: "br-ex"                                          |
  42 |              | attacker2(for management netork):                            |
  43 |              | -fault_type: "close-interface"                               |
  44 |              | -host: node1                                                 |
  45 |              | -interface: "br-mgmt"                                        |
  46 |              | attacker3(for storage netork):                               |
  47 |              | -fault_type: "close-interface"                               |
  48 |              | -host: node1                                                 |
  49 |              | -interface: "br-storage"                                     |
  50 |              | attacker4(for private netork):                               |
  51 |              | -fault_type: "close-interface"                               |
  52 |              | -host: node1                                                 |
  53 |              | -interface: "br-mesh"                                        |
  54 +--------------+--------------------------------------------------------------+
  55 |monitors      | In this test case, the monitor named "openstack-cmd" is      |
  56 |              | needed. The monitor needs needs two parameters:              |
  57 |              | 1) monitor_type: which is used for finding the monitor class |
  58 |              | and related scritps. It should be always set to              |
  59 |              | "openstack-cmd" for this monitor.                            |
  60 |              | 2) command_name: which is the command name used for request  |
  61 |              |                                                              |
  62 |              | There are four instance of the "openstack-cmd" monitor:      |
  63 |              | monitor1:                                                    |
  64 |              | -monitor_type: "openstack-cmd"                               |
  65 |              | -command_name: "nova image-list"                             |
  66 |              | monitor2:                                                    |
  67 |              | -monitor_type: "openstack-cmd"                               |
  68 |              | -command_name: "neutron router-list"                         |
  69 |              | monitor3:                                                    |
  70 |              | -monitor_type: "openstack-cmd"                               |
  71 |              | -command_name: "heat stack-list"                             |
  72 |              | monitor4:                                                    |
  73 |              | -monitor_type: "openstack-cmd"                               |
  74 |              | -command_name: "cinder list"                                 |
  75 +--------------+--------------------------------------------------------------+
  76 |metrics       | In this test case, there is one metric:                      |
  77 |              | 1)service_outage_time: which indicates the maximum outage    |
  78 |              | time (seconds) of the specified Openstack command request.   |
  79 +--------------+--------------------------------------------------------------+
  80 |test tool     | Developed by the project. Please see folder:                 |
  81 |              | "yardstick/benchmark/scenarios/availability/ha_tools"        |
  82 |              |                                                              |
  83 +--------------+--------------------------------------------------------------+
  84 |references    | ETSI NFV REL001                                              |
  85 |              |                                                              |
  86 +--------------+--------------------------------------------------------------+
  87 |configuration | This test case needs two configuration files:                |
  88 |              | 1) test case file: opnfv_yardstick_tc050.yaml                |
  89 |              | -Attackers: see above "attackers" discription                |
  90 |              | -waiting_time: which is the time (seconds) from the process  |
  91 |              | being killed to stoping monitors the monitors                |
  92 |              | -Monitors: see above "monitors" discription                  |
  93 |              | -SLA: see above "metrics" discription                        |
  94 |              |                                                              |
  95 |              | 2)POD file: pod.yaml                                         |
  96 |              | The POD configuration should record on pod.yaml first.       |
  97 |              | the "host" item in this test case will use the node name in  |
  98 |              | the pod.yaml.                                                |
  99 |              |                                                              |
 100 +--------------+--------------------------------------------------------------+
 101 |test sequence | description and expected result                              |
 102 |              |                                                              |
 103 +--------------+--------------------------------------------------------------+
 104 |step 1        | start monitors:                                              |
 105 |              | each monitor will run with independently process             |
 106 |              |                                                              |
 107 |              | Result: The monitor info will be collected.                  |
 108 |              |                                                              |
 109 +--------------+--------------------------------------------------------------+
 110 |step 2        | do attacker: connect the host through SSH, and then execute  |
 111 |              | the turnoff network interface script with param value        |
 112 |              | specified by  "interface".                                   |
 113 |              |                                                              |
 114 |              | Result: Network interfaces will be turned down.              |
 115 |              |                                                              |
 116 +--------------+--------------------------------------------------------------+
 117 |step 3        | stop monitors after a period of time specified by            |
 118 |              | "waiting_time"                                               |
 119 |              |                                                              |
 120 |              | Result: The monitor info will be aggregated.                 |
 121 |              |                                                              |
 122 +--------------+--------------------------------------------------------------+
 123 |step 4        | verify the SLA                                               |
 124 |              |                                                              |
 125 |              | Result: The test case is passed or not.                      |
 126 |              |                                                              |
 127 +--------------+--------------------------------------------------------------+
 128 |post-action   | It is the action when the test cases exist. It turns up the  |
 129 |              | network interface of the control node if it is not turned    |
 130 |              | up.                                                          |
 131 +--------------+--------------------------------------------------------------+
 132 |test verdict  | Fails only if SLA is not passed, or if there is a test case  |
 133 |              | execution problem.                                           |
 134 |              |                                                              |
 135 +--------------+--------------------------------------------------------------+