1 .. This work is licensed under a Creative Commons Attribution 4.0 International
3 .. http://creativecommons.org/licenses/by/4.0
4 .. (c) OPNFV, Yin Kanglin and others.
5 .. 14_ykl@tongji.edu.cn
7 *************************************
8 Yardstick Test Case Description TC046
9 *************************************
11 +-----------------------------------------------------------------------------+
12 |Control Node Openstack Service High Availability - Keystone |
14 +--------------+--------------------------------------------------------------+
15 |test case id | OPNFV_YARDSTICK_TC046: Control node Openstack service down - |
17 +--------------+--------------------------------------------------------------+
18 |test purpose | This test case will verify the high availability of the |
19 | | user service provided by OpenStack (keystone) on control |
22 +--------------+--------------------------------------------------------------+
23 |test method | This test case kills the processes of keystone service on a |
24 | | selected control node, then checks whether the request of |
25 | | the related Openstack command is OK and the killed processes |
28 +--------------+--------------------------------------------------------------+
29 |attackers | In this test case, an attacker called "kill-process" 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 "kill-process" in this |
34 | | 2) process_name: which is the process name of the specified |
35 | | OpenStack service. If there are multiple processes use the |
36 | | same name on the host, all of them are killed by this |
38 | | In this case. This parameter should always set to "keystone" |
39 | | 3) host: which is the name of a control node being attacked. |
42 | | -fault_type: "kill-process" |
43 | | -process_name: "keystone" |
46 +--------------+--------------------------------------------------------------+
47 |monitors | In this test case, two kinds of monitor are needed: |
48 | | 1. the "openstack-cmd" monitor constantly request a specific |
49 | | Openstack command, which needs two parameters: |
50 | | 1) monitor_type: which is used for finding the monitor class |
51 | | and related scritps. It should be always set to |
52 | | "openstack-cmd" for this monitor. |
53 | | 2) command_name: which is the command name used for request. |
54 | | In this case, the command name should be keystone related |
57 | | 2. the "process" monitor check whether a process is running |
58 | | on a specific node, which needs three parameters: |
59 | | 1) monitor_type: which used for finding the monitor class and|
60 | | related scritps. It should be always set to "process" |
61 | | for this monitor. |
62 | | 2) process_name: which is the process name for monitor |
63 | | 3) host: which is the name of the node runing the process |
67 | | -monitor_type: "openstack-cmd" |
68 | | -command_name: "keystone user-list" |
70 | | -monitor_type: "process" |
71 | | -process_name: "keystone" |
74 +--------------+--------------------------------------------------------------+
75 |metrics | In this test case, there are two metrics: |
76 | | 1)service_outage_time: which indicates the maximum outage |
77 | | time (seconds) of the specified Openstack command request. |
78 | | 2)process_recover_time: which indicates the maximun time |
79 | | (seconds) from the process being killed to recovered |
81 +--------------+--------------------------------------------------------------+
82 |test tool | Developed by the project. Please see folder: |
83 | | "yardstick/benchmark/scenarios/availability/ha_tools" |
85 +--------------+--------------------------------------------------------------+
86 |references | ETSI NFV REL001 |
88 +--------------+--------------------------------------------------------------+
89 |configuration | This test case needs two configuration files: |
90 | | 1) test case file: opnfv_yardstick_tc046.yaml |
91 | | -Attackers: see above "attackers" discription |
92 | | -waiting_time: which is the time (seconds) from the process |
93 | | being killed to stoping monitors the monitors |
94 | | -Monitors: see above "monitors" discription |
95 | | -SLA: see above "metrics" discription |
97 | | 2)POD file: pod.yaml |
98 | | The POD configuration should record on pod.yaml first. |
99 | | the "host" item in this test case will use the node name in |
102 +--------------+--------------------------------------------------------------+
103 |test sequence | description and expected result |
105 +--------------+--------------------------------------------------------------+
106 |step 1 | start monitors: |
107 | | each monitor will run with independently process |
109 | | Result: The monitor info will be collected. |
111 +--------------+--------------------------------------------------------------+
112 |step 2 | do attacker: connect the host through SSH, and then execute |
113 | | the kill process script with param value specified by |
116 | | Result: Process will be killed. |
118 +--------------+--------------------------------------------------------------+
119 |step 3 | stop monitors after a period of time specified by |
122 | | Result: The monitor info will be aggregated. |
124 +--------------+--------------------------------------------------------------+
125 |step 4 | verify the SLA |
127 | | Result: The test case is passed or not. |
129 +--------------+--------------------------------------------------------------+
130 |post-action | It is the action when the test cases exist. It will check |
131 | | the status of the specified process on the host, and restart |
132 | | the process if it is not running for next test cases. |
134 | | Notice: This post-action uses 'lsb_release' command to check |
135 | | the host linux distribution and determine the OpenStack |
136 | | service name to restart the process. Lack of 'lsb_release' |
137 | | on the host may cause failure to restart the process. |
139 +--------------+--------------------------------------------------------------+
140 |test verdict | Fails only if SLA is not passed, or if there is a test case |
141 | | execution problem. |
143 +--------------+--------------------------------------------------------------+