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
|
/* Copyright (c) 2017-2025. The SimGrid Team. All rights reserved. */
/* This program is free software; you can redistribute it and/or modify it
* under the terms of the license (GNU LGPL) which comes with this package. */
/* This example illustrates how to use Simgrid Tasks to reproduce the workflow 2.a of the article
* "Automated performance prediction of microservice applications using simulation" by Clément Courageux-Sudan et al.
*
* To build this workflow we create;
* - each Execution Task
* - each Communication Task
* - the links between the Tasks, i.e. the graph
*
* We also increase the parallelism degree of each Task to 10.
*
* In this scenario we send 500 requests per second (RPS) to the entry point of the graph for 7 seconds,
* and we count the number of processed requests between 2 and 7 seconds to evaluate the number of requests processed
* per second.
*/
#include "simgrid/s4u.hpp"
XBT_LOG_NEW_DEFAULT_CATEGORY(task_microservice, "Messages specific for this s4u example");
namespace sg4 = simgrid::s4u;
static void request_sender(sg4::TaskPtr t, int requests_per_second)
{
for (int i = 0; i < requests_per_second * 7; i++) {
t->enqueue_firings(1);
sg4::this_actor::sleep_for(1.0 / requests_per_second);
}
}
int main(int argc, char* argv[])
{
sg4::Engine e(&argc, argv);
e.load_platform(argv[1]);
// Retrieve Hosts
auto pm0 = e.host_by_name("PM0");
auto pm1 = e.host_by_name("PM1");
// Set concurrency limit
pm0->set_concurrency_limit(10);
pm1->set_concurrency_limit(10);
// Create Exec Tasks
auto nginx_web_server = sg4::ExecTask::init("nginx_web_server", 783 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_0 = sg4::ExecTask::init("compose_post_service_0", 682 * pm0->get_speed() / 1e6, pm0);
auto unique_id_service = sg4::ExecTask::init("unique_id_service", 12 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_1 = sg4::ExecTask::init("compose_post_service_1", 140 * pm0->get_speed() / 1e6, pm0);
auto media_service = sg4::ExecTask::init("media_service", 6 * pm1->get_speed() / 1e6, pm1);
auto compose_post_service_2 = sg4::ExecTask::init("compose_post_service_2", 135 * pm0->get_speed() / 1e6, pm0);
auto user_service = sg4::ExecTask::init("user_service", 5 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_3 = sg4::ExecTask::init("compose_post_service_3", 147 * pm0->get_speed() / 1e6, pm0);
auto text_service_0 = sg4::ExecTask::init("text_service_0", 296 * pm0->get_speed() / 1e6, pm0);
auto text_service_1 = sg4::ExecTask::init("text_service_1", 350 * pm0->get_speed() / 1e6, pm0);
auto text_service_2 = sg4::ExecTask::init("text_service_2", 146 * pm0->get_speed() / 1e6, pm0);
auto user_mention_service = sg4::ExecTask::init("user_mention_service", 934 * pm0->get_speed() / 1e6, pm0);
auto url_shorten_service = sg4::ExecTask::init("url_shorten_service", 555 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_4 = sg4::ExecTask::init("compose_post_service_4", 138 * pm0->get_speed() / 1e6, pm0);
auto home_timeline_service_0 = sg4::ExecTask::init("home_timeline_service_0", 243 * pm0->get_speed() / 1e6, pm0);
auto social_graph_service = sg4::ExecTask::init("home_timeline_service_0", 707 * pm0->get_speed() / 1e6, pm0);
auto home_timeline_service_1 = sg4::ExecTask::init("home_timeline_service_0", 7 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_5 = sg4::ExecTask::init("compose_post_service_5", 192 * pm0->get_speed() / 1e6, pm0);
auto user_timeline_service = sg4::ExecTask::init("user_timeline_service", 913 * pm0->get_speed() / 1e6, pm0);
auto compose_post_service_6 = sg4::ExecTask::init("compose_post_service_6", 508 * pm0->get_speed() / 1e6, pm0);
auto post_storage_service = sg4::ExecTask::init("post_storage_service", 391 * pm1->get_speed() / 1e6, pm1);
// Create Comm Tasks
auto compose_post_service_1_to_media_service =
sg4::CommTask::init("compose_post_service_1_to_media_service", 100, pm0, pm1);
auto media_service_to_compose_post_service_2 =
sg4::CommTask::init("media_service_to_compose_post_service_2", 100, pm1, pm0);
auto compose_post_service_6_to_post_storage_service =
sg4::CommTask::init("media_service_to_compose_post_service_2", 100, pm1, pm0);
// Create the graph
nginx_web_server->add_successor(compose_post_service_0);
compose_post_service_0->add_successor(unique_id_service);
unique_id_service->add_successor(compose_post_service_1);
compose_post_service_1->add_successor(compose_post_service_1_to_media_service);
compose_post_service_1_to_media_service->add_successor(media_service);
media_service->add_successor(media_service_to_compose_post_service_2);
media_service_to_compose_post_service_2->add_successor(compose_post_service_2);
compose_post_service_2->add_successor(user_service);
user_service->add_successor(compose_post_service_3);
compose_post_service_3->add_successor(text_service_0);
text_service_0->add_successor(text_service_1);
text_service_0->add_successor(text_service_2);
text_service_1->add_successor(user_mention_service);
text_service_2->add_successor(url_shorten_service);
user_mention_service->add_successor(compose_post_service_4);
compose_post_service_4->add_successor(home_timeline_service_0);
home_timeline_service_0->add_successor(social_graph_service);
social_graph_service->add_successor(home_timeline_service_1);
home_timeline_service_1->add_successor(compose_post_service_5);
compose_post_service_5->add_successor(user_timeline_service);
user_timeline_service->add_successor(compose_post_service_6);
compose_post_service_6->add_successor(compose_post_service_6_to_post_storage_service);
compose_post_service_6_to_post_storage_service->add_successor(post_storage_service);
// Dispatch Exec Tasks
std::vector<sg4::TaskPtr> exec_tasks = {nginx_web_server,
compose_post_service_0,
unique_id_service,
compose_post_service_1,
compose_post_service_1_to_media_service,
media_service,
media_service_to_compose_post_service_2,
compose_post_service_2,
user_service,
compose_post_service_3,
text_service_0,
text_service_1,
text_service_2,
user_mention_service,
url_shorten_service,
compose_post_service_4,
home_timeline_service_0,
social_graph_service,
home_timeline_service_1,
compose_post_service_5,
user_timeline_service,
compose_post_service_6,
compose_post_service_6_to_post_storage_service,
post_storage_service};
for (auto t : exec_tasks)
t->set_parallelism_degree(10);
// Create the actor that will inject requests during the simulation
e.add_actor("request_sender", pm0, request_sender, nginx_web_server, 500);
// Add a function to be called when tasks end for log purpose
int requests_processed = 0;
sg4::Task::on_completion_cb([&requests_processed](const sg4::Task* t) {
auto timenow = sg4::Engine::get_clock();
if (t->get_name() == "post_storage_service" and timenow < 7 and timenow > 2)
requests_processed++;
});
// Start the simulation
e.run();
XBT_INFO("Requests processed per second: %f", requests_processed / 5.0);
return 0;
}
|
