1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
|
@sbd
Feature: configure sbd delay start correctly
Tag @clean means need to stop cluster service if the service is available
@clean
Scenario: disk-based SBD with small sbd_watchdog_timeout
Given Run "test -f /etc/crm/profiles.yml" OK
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "15"
Given Has disk "/dev/sda1" on "hanode1"
Given Cluster service is "stopped" on "hanode1"
When Run "crm cluster init -s /dev/sda1 -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
And Service "sbd" is "started" on "hanode1"
And Resource "stonith-sbd" type "fence_sbd" is "Started"
And SBD option "SBD_DELAY_START" value is "no"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
# original value is 43, which is calculated by external/sbd RA
# now fence_sbd doesn't calculate it, so this value is the default one
# from pacemaker
And Cluster property "stonith-timeout" is "60"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
Given Has disk "/dev/sda1" on "hanode2"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Service "sbd" is "started" on "hanode2"
# SBD_DELAY_START >= (token + consensus + pcmk_delay_max + msgwait) # for disk-based sbd
And SBD option "SBD_DELAY_START" value is "71"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
# value_from_sbd >= 1.2 * msgwait # for disk-based sbd
# stonith_timeout >= max(value_from_sbd, constants.STONITH_TIMEOUT_DEFAULT) + token + consensus
And Cluster property "stonith-timeout" is "71"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
Given Has disk "/dev/sda1" on "hanode3"
Given Cluster service is "stopped" on "hanode3"
When Run "crm cluster join -c hanode1 -y" on "hanode3"
Then Cluster service is "started" on "hanode3"
And Service "sbd" is "started" on "hanode3"
# SBD_DELAY_START >= (token + consensus + pcmk_delay_max + msgwait) # for disk-based sbd
# runtime value is "41", we keep the larger one here
And SBD option "SBD_DELAY_START" value is "71"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
# value_from_sbd >= 1.2 * msgwait # for disk-based sbd
# stonith_timeout >= max(value_from_sbd, constants.STONITH_TIMEOUT_DEFAULT) + token + consensus
# runtime value is "71", we keep ther larger one here
And Cluster property "stonith-timeout" is "71"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
When Run "crm cluster remove hanode3 -y" on "hanode1"
Then Cluster service is "stopped" on "hanode3"
And Service "sbd" is "stopped" on "hanode3"
And SBD option "SBD_DELAY_START" value is "71"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
And Cluster property "stonith-timeout" is "71"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
@clean
Scenario: disk-less SBD with small sbd_watchdog_timeout
Given Run "test -f /etc/crm/profiles.yml" OK
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "15"
Given Cluster service is "stopped" on "hanode1"
When Run "crm cluster init -S -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
And SBD option "SBD_DELAY_START" value is "no"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "15"
And Cluster property "stonith-timeout" is "60"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
# SBD_DELAY_START >= (token + consensus + 2*SBD_WATCHDOG_TIMEOUT) # for disk-less sbd
And SBD option "SBD_DELAY_START" value is "41"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "15"
# stonith-timeout >= 1.2 * max(stonith_watchdog_timeout, 2*SBD_WATCHDOG_TIMEOUT) # for disk-less sbd
# stonith_timeout >= max(value_from_sbd, constants.STONITH_TIMEOUT_DEFAULT) + token + consensus
And Cluster property "stonith-timeout" is "71"
Given Cluster service is "stopped" on "hanode3"
When Run "crm cluster join -c hanode1 -y" on "hanode3"
Then Cluster service is "started" on "hanode3"
And SBD option "SBD_DELAY_START" value is "41"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "15"
And Cluster property "stonith-timeout" is "71"
When Run "crm cluster remove hanode3 -y" on "hanode1"
Then Cluster service is "stopped" on "hanode3"
And SBD option "SBD_DELAY_START" value is "41"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "15"
And Cluster property "stonith-timeout" is "71"
@clean
Scenario: disk-based SBD with big sbd_watchdog_timeout
When Run "sed -i 's/watchdog_timeout: 15/watchdog_timeout: 60/' /etc/crm/profiles.yml" on "hanode1"
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "60"
Given Has disk "/dev/sda1" on "hanode1"
Given Cluster service is "stopped" on "hanode1"
When Run "crm cluster init -s /dev/sda1 -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
And Service "sbd" is "started" on "hanode1"
And Resource "stonith-sbd" type "fence_sbd" is "Started"
And SBD option "SBD_DELAY_START" value is "no"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "120"
# original value is 172, which is calculated by external/sbd RA
# now fence_sbd doesn't calculate it, so this value is the default one
# from pacemaker
And Cluster property "stonith-timeout" is "60"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
Given Has disk "/dev/sda1" on "hanode2"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Service "sbd" is "started" on "hanode2"
# SBD_DELAY_START >= (token + consensus + pcmk_delay_max + msgwait) # for disk-based sbd
And SBD option "SBD_DELAY_START" value is "161"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "120"
# stonith-timeout >= 1.2 * msgwait # for disk-based sbd
# stonith_timeout >= max(value_from_sbd, constants.STONITH_TIMEOUT_DEFAULT) + token + consensus
And Cluster property "stonith-timeout" is "155"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
# since SBD_DELAY_START value(161s) > default systemd startup value(1min 30s)
And Run "test -f /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
# 1.2*SBD_DELAY_START
And Run "grep 'TimeoutSec=193' /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
Given Has disk "/dev/sda1" on "hanode3"
Given Cluster service is "stopped" on "hanode3"
When Run "crm cluster join -c hanode1 -y" on "hanode3"
Then Cluster service is "started" on "hanode3"
And Service "sbd" is "started" on "hanode3"
And SBD option "SBD_DELAY_START" value is "161"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "120"
And Cluster property "stonith-timeout" is "155"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
And Run "test -f /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
And Run "grep 'TimeoutSec=193' /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
When Run "crm cluster remove hanode3 -y" on "hanode1"
Then Cluster service is "stopped" on "hanode3"
And Service "sbd" is "stopped" on "hanode3"
And SBD option "SBD_DELAY_START" value is "161"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "120"
And Cluster property "stonith-timeout" is "155"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
And Run "test -f /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
And Run "grep 'TimeoutSec=193' /etc/systemd/system/sbd.service.d/sbd_delay_start.conf" OK
When Run "sed -i 's/watchdog_timeout: 60/watchdog_timeout: 15/g' /etc/crm/profiles.yml" on "hanode1"
@clean
Scenario: Add sbd via stage on a running cluster
Given Run "test -f /etc/crm/profiles.yml" OK
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "15"
Given Has disk "/dev/sda1" on "hanode1"
Given Has disk "/dev/sda1" on "hanode2"
Given Cluster service is "stopped" on "hanode1"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster init -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
When Run "crm cluster init sbd -s /dev/sda1 -y" on "hanode1"
Then Service "sbd" is "started" on "hanode1"
Then Service "sbd" is "started" on "hanode2"
And SBD option "SBD_DELAY_START" value is "71"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
And Cluster property "stonith-timeout" is "71"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
@clean
Scenario: Add disk-based sbd with qdevice
Given Run "test -f /etc/crm/profiles.yml" OK
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "15"
Given Has disk "/dev/sda1" on "hanode1"
Given Has disk "/dev/sda1" on "hanode2"
Given Cluster service is "stopped" on "hanode1"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster init -s /dev/sda1 --qnetd-hostname=qnetd-node -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Service "corosync-qdevice" is "started" on "hanode1"
And Service "corosync-qdevice" is "started" on "hanode2"
And Service "sbd" is "started" on "hanode1"
And Service "sbd" is "started" on "hanode2"
And SBD option "SBD_DELAY_START" value is "41"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "5"
And SBD option "msgwait" value for "/dev/sda1" is "30"
And Cluster property "stonith-timeout" is "71"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
@clean
Scenario: Add disk-less sbd with qdevice
Given Run "test -f /etc/crm/profiles.yml" OK
Given Yaml "default:corosync.totem.token" value is "5000"
Given Yaml "default:sbd.watchdog_timeout" value is "15"
Given Cluster service is "stopped" on "hanode1"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster init -S --qnetd-hostname=qnetd-node -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Service "corosync-qdevice" is "started" on "hanode1"
And Service "corosync-qdevice" is "started" on "hanode2"
And Service "sbd" is "started" on "hanode1"
And Service "sbd" is "started" on "hanode2"
And SBD option "SBD_DELAY_START" value is "81"
And SBD option "SBD_WATCHDOG_TIMEOUT" value is "35"
And Cluster property "stonith-timeout" is "95"
And Cluster property "stonith-watchdog-timeout" is "70"
@clean
Scenario: Add and remove qdevice from cluster with sbd running
Given Cluster service is "stopped" on "hanode1"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster init -s /dev/sda1 -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Service "sbd" is "started" on "hanode1"
And Service "sbd" is "started" on "hanode2"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
When Run "crm cluster init qdevice --qnetd-hostname=qnetd-node -y" on "hanode1"
Then Service "corosync-qdevice" is "started" on "hanode1"
And Service "corosync-qdevice" is "started" on "hanode2"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
When Run "crm cluster remove --qdevice -y" on "hanode1"
Then Service "corosync-qdevice" is "stopped" on "hanode1"
And Service "corosync-qdevice" is "stopped" on "hanode2"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
@clean
Scenario: Test priority-fence-delay and priority
Given Cluster service is "stopped" on "hanode1"
Given Cluster service is "stopped" on "hanode2"
When Run "crm cluster init -y" on "hanode1"
Then Cluster service is "started" on "hanode1"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Property "priority" in "rsc_defaults" is "1"
When Run "crm cluster remove hanode2 -y" on "hanode1"
Then Cluster service is "stopped" on "hanode2"
And Property "priority" in "rsc_defaults" is "0"
When Run "crm cluster join -c hanode1 -y" on "hanode2"
Then Cluster service is "started" on "hanode2"
And Property "priority" in "rsc_defaults" is "1"
When Run "crm cluster init qdevice --qnetd-hostname=qnetd-node -y" on "hanode1"
Then Service "corosync-qdevice" is "started" on "hanode1"
And Service "corosync-qdevice" is "started" on "hanode2"
And Property "priority" in "rsc_defaults" is "0"
When Run "crm cluster remove --qdevice -y" on "hanode1"
Then Service "corosync-qdevice" is "stopped" on "hanode1"
And Service "corosync-qdevice" is "stopped" on "hanode2"
And Property "priority" in "rsc_defaults" is "1"
When Run "crm cluster init sbd -s /dev/sda1 -y" on "hanode1"
Then Service "sbd" is "started" on "hanode1"
And Service "sbd" is "started" on "hanode2"
And Parameter "pcmk_delay_max" configured in "stonith-sbd"
And Cluster property "stonith-timeout" is "71"
And Cluster property "priority-fencing-delay" is "60"
When Run "crm cluster remove hanode2 -y" on "hanode1"
Then Cluster service is "stopped" on "hanode2"
And Property "priority" in "rsc_defaults" is "0"
And Cluster property "priority-fencing-delay" is "0"
And Parameter "pcmk_delay_max" not configured in "stonith-sbd"
|
