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 TC058
9 *************************************
11 +-----------------------------------------------------------------------------+
12 |OpenStack Controller Virtual Router Service High Availability |
13 +==============+==============================================================+
14 |test case id | OPNFV_YARDSTICK_TC058:OpenStack Controller Virtual Router |
15 | | Service High Availability |
16 +--------------+--------------------------------------------------------------+
17 |test purpose | This test case will verify the high availability of virtual |
18 | | routers(L3 agent) on controller node. When a virtual router |
19 | | service on a specified controller node is shut down, this |
20 | | test case will check whether the network of virtual machines |
21 | | will be affected, and whether the attacked virtual router |
22 | | service will be recovered. |
23 +--------------+--------------------------------------------------------------+
24 |test method | This test case kills the processes of virtual router service |
25 | | (l3-agent) on a selected controller node(the node holds the |
26 | | active l3-agent), then checks whether the network routing |
27 | | of virtual machines is OK and whether the killed service |
28 | | will be recovered. |
29 +--------------+--------------------------------------------------------------+
30 |attackers | In this test case, an attacker called "kill-process" is |
31 | | needed. This attacker includes three parameters: |
32 | | 1) fault_type: which is used for finding the attacker's |
33 | | scripts. It should be always set to "kill-process" in this |
35 | | 2) process_name: which is the process name of the load |
36 | | balance service. If there are multiple processes use the |
37 | | same name on the host, all of them are killed by this |
39 | | 3) host: which is the name of a control node being attacked. |
41 | | In this case, this process name should set to "l3agent" , |
43 | | -fault_type: "kill-process" |
44 | | -process_name: "l3agent" |
46 +--------------+--------------------------------------------------------------+
47 |monitors | In this test case, two kinds of monitor are needed: |
48 | | 1. the "ip_status" monitor that pings a specific ip to check |
49 | | the connectivity of this ip, which needs two parameters: |
50 | | 1) monitor_type: which is used for finding the monitor class |
51 | | and related scripts. It should be always set to "ip_status" |
52 | | for this monitor. |
53 | | 2) ip_address: The ip to be pinged. In this case, ip_address |
54 | | will be either an ip address of external network or an ip |
55 | | address of a virtual machine. |
56 | | 3) host: The node on which ping will be executed, in this |
57 | | case the host will be a virtual machine. |
59 | | 2. the "process" monitor check whether a process is running |
60 | | on a specific node, which needs three parameters: |
61 | | 1) monitor_type: which used for finding the monitor class |
62 | | and related scripts. It should be always set to "process" |
63 | | for this monitor. |
64 | | 2) process_name: which is the process name for monitor. In |
65 | | this case, the process-name of monitor2 should be "l3agent" |
66 | | 3) host: which is the name of the node running the process |
70 | | -monitor_type: "ip_status" |
71 | | -host: 172.16.0.11 |
72 | | -ip_address: 172.16.1.11 |
74 | | -monitor_type: "ip_status" |
75 | | -host: 172.16.0.11 |
76 | | -ip_address: 8.8.8.8 |
78 | | -monitor_type: "process" |
79 | | -process_name: "l3agent" |
81 +--------------+--------------------------------------------------------------+
82 |metrics | In this test case, there are two metrics: |
83 | | 1)service_outage_time: which indicates the maximum outage |
84 | | time (seconds) of the specified Openstack command request. |
85 | | 2)process_recover_time: which indicates the maximum time |
86 | | (seconds) from the process being killed to recovered |
87 +--------------+--------------------------------------------------------------+
88 |test tool | None. Self-developed. |
89 +--------------+--------------------------------------------------------------+
90 |references | ETSI NFV REL001 |
91 +--------------+--------------------------------------------------------------+
92 |configuration | This test case needs two configuration files: |
93 | | 1) test case file: opnfv_yardstick_tc058.yaml |
94 | | -Attackers: see above "attackers" description |
95 | | -Monitors: see above "monitors" description |
96 | | -Steps: the test case execution step, see "test sequence" |
97 | | description below |
99 | | 2)POD file: pod.yaml |
100 | | The POD configuration should record on pod.yaml first. |
101 | | the "host" item in this test case will use the node name in |
103 +--------------+------+----------------------------------+--------------------+
104 |test sequence | description and expected result |
106 +--------------+--------------------------------------------------------------+
107 |pre-test | The test case image needs to be installed into Glance |
108 |conditions | with cachestat included in the image. |
110 +--------------+--------------------------------------------------------------+
111 |step 1 | Two host VMs are booted, these two hosts are in two different|
112 | | networks, the networks are connected by a virtual router |
114 +--------------+--------------------------------------------------------------+
115 |step 1 | start monitors: |
116 | | each monitor will run with independently process |
118 | | Result: The monitor info will be collected. |
120 +--------------+--------------------------------------------------------------+
121 |step 2 | do attacker: connect the host through SSH, and then execute |
122 | | the kill process script with param value specified by |
125 | | Result: Process will be killed. |
127 +--------------+--------------------------------------------------------------+
128 |step 4 | stop monitors after a period of time specified by |
131 | | Result: The monitor info will be aggregated. |
133 +--------------+--------------------------------------------------------------+
134 |step 5 | verify the SLA |
136 | | Result: The test case is passed or not. |
138 +--------------+------+----------------------------------+--------------------+
139 |post-action | It is the action when the test cases exist. It will check |
140 | | the status of the specified process on the host, and restart |
141 | | the process if it is not running for next test cases. |
142 | | Virtual machines and network created in the test case will |
145 +--------------+------+----------------------------------+--------------------+
146 |test verdict | Fails only if SLA is not passed, or if there is a test case |
147 | | execution problem. |
148 +--------------+--------------------------------------------------------------+