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
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
|
# MultiNode job definition
A LAVA MultiNode job is a single LAVA job definition that runs multiple jobs
across multiple devices. It is implemented via the
[MultiNode](../../technical-references/job-definition/protocols.md#multinode)
protocol that allows the jobs to communicate via the protocol API and requests.
## How to avoid it
While the MultiNode support in LAVA is powerful and mature, it also introduces
significant complexity for writing and debugging the job definitions. We
recommend avoiding MultiNode unless it is truly required by your use case.
Since the introduction of the [docker test shell](./running-arbitrary-code-with-docker.md)
for single node jobs, MultiNode is no longer required in many scenarios.
Communicating with a DUT via a Docker container within a single node job is a
much simpler and more reliable approach. Give it a try before you start working
with the MultiNode.
## Dependency
[LAVA coordinator](../../technical-references/services/lava-coordinator.md)
service should be configured and reachable from all the workers involved in the
MultiNode jobs.
## Coordinating via APIs
The following example job uses the
[MultiNode APIs](../../technical-references/job-definition/protocols.md#api)
from the LAVA test shell to pass messages between the iPerf server and client for
starting the test.
```yaml hl_lines="71 72 125 128"
job_name: MultiNode iPerf test
visibility: public
priority: medium
timeouts:
job:
minutes: 30
action:
minutes: 5
connection:
minutes: 2
actions:
power-off:
seconds: 30
# Define roles
protocols:
lava-multinode:
roles:
server:
device_type: ssh
count: 1
client:
device_type: bcm2711-rpi-4-b
count: 1
timeout:
minutes: 10
actions:
# Server actions
- deploy:
role:
- server
to: ssh
os: debian
timeout:
minutes: 5
- boot:
role:
- server
method: ssh
prompts:
- 'root@x13:~#'
timeout:
minutes: 5
- test:
role:
- server
protocols:
lava-multinode:
- action: multinode-test
request: lava-wait
messageID: start
timeout:
minutes: 10
definitions:
- from: inline
repository:
metadata:
format: Lava-Test Test Definition 1.0
name: iperf-server
run:
steps:
- iperf -s &
- echo $! > /tmp/iperf-server.pid
- IP=$(ip route get 8.8.8.8 | grep -oP 'src \K\S+')
- lava-test-case get-server-ip --shell echo "Server IP is ${IP}"
- lava-send server-ready server_ip=${IP}
- lava-wait client-done
- kill -9 $(cat /tmp/iperf-server.pid)
name: iperf-server
path: inline/iperf-server.yaml
timeout:
minutes: 15
# Client actions
- deploy:
role:
- client
to: usbg-ms
image:
url: http://198.18.0.1/tmp/images/rpi4/20231109_raspi_4_bookworm.img
format: ext4
partition: 1
overlays:
lava: true
timeout:
minutes: 5
- boot:
role:
- client
method: minimal
auto_login:
login_prompt: 'login:'
username: root
prompts:
- 'root@rpi4-20231108:'
timeout:
minutes: 5
- test:
protocols:
lava-multinode:
- action: multinode-test
request: lava-send
messageID: start
timeout:
minutes: 10
role:
- client
definitions:
- from: inline
repository:
metadata:
format: Lava-Test Test Definition 1.0
name: iperf-client
run:
steps:
- apt-get update -q
- lava-test-case install-iperf --shell apt-get -q -y install iperf
- lava-wait server-ready
- SERVER_IP=$(cat /tmp/lava_multi_node_cache.txt | cut -d = -f 2)
- lava-test-case iperf-to-server --shell iperf -c ${SERVER_IP} -t 10
- lava-send client-done
name: iperf-client
path: inline/iperf-client.yaml
timeout:
minutes: 15
```
## Coordinating via requests
The following example job uses the
[MultiNode protocol requests](../../technical-references/job-definition/protocols.md#request)
across multiple job actions to configure the faster server waits for the slower
client.
```yaml hl_lines="36-42 109-116"
job_name: MultiNode iPerf test - wait for client
visibility: public
priority: medium
timeouts:
job:
minutes: 30
action:
minutes: 5
connection:
minutes: 2
actions:
power-off:
seconds: 30
# Define roles
protocols:
lava-multinode:
roles:
server:
device_type: ssh
count: 1
client:
device_type: bcm2711-rpi-4-b
count: 1
timeout:
minutes: 10
actions:
# Server actions
- deploy:
role:
- server
# Wait for the client.
protocols:
lava-multinode:
- action: scp-overlay
request: lava-wait
messageID: start
timeout:
minutes: 10
to: ssh
os: debian
timeout:
minutes: 5
- boot:
role:
- server
method: ssh
prompts:
- 'root@x13:~#'
timeout:
minutes: 5
- test:
role:
- server
definitions:
- from: inline
repository:
metadata:
format: Lava-Test Test Definition 1.0
name: iperf-server
run:
steps:
- iperf -s &
- echo $! > /tmp/iperf-server.pid
- IP=$(ip route get 8.8.8.8 | grep -oP 'src \K\S+')
- lava-test-case get-server-ip --shell echo "Server IP is ${IP}"
- lava-send server-ready server_ip=${IP}
- lava-wait client-done
- kill -9 $(cat /tmp/iperf-server.pid)
name: iperf-server
path: inline/iperf-server.yaml
timeout:
minutes: 15
# Client actions
- deploy:
role:
- client
to: usbg-ms
image:
url: http://198.18.0.1/tmp/images/rpi4/20231109_raspi_4_bookworm.img
format: ext4
partition: 1
overlays:
lava: true
timeout:
minutes: 5
- boot:
role:
- client
method: minimal
auto_login:
login_prompt: 'login:'
username: root
prompts:
- 'root@rpi4-20231108:'
timeout:
minutes: 5
- test:
role:
- client
protocols:
lava-multinode:
- action: multinode-test
# Notify sever to start.
request: lava-send
messageID: start
timeout:
minutes: 10
definitions:
- from: inline
repository:
metadata:
format: Lava-Test Test Definition 1.0
name: iperf-client
run:
steps:
- apt-get update -q
- lava-test-case install-iperf --shell apt-get -q -y install iperf
- lava-wait server-ready
- SERVER_IP=$(cat /tmp/lava_multi_node_cache.txt | cut -d = -f 2)
- lava-test-case iperf-to-server --shell iperf -c ${SERVER_IP} -t 10
- lava-send client-done
name: iperf-client
path: inline/iperf-client.yaml
timeout:
minutes: 15
```
|
