1 /* Copyright (c) 2021-2023. The SimGrid Team. All rights reserved. */
3 /* This program is free software; you can redistribute it and/or modify it
4 * under the terms of the license (GNU LGPL) which comes with this package. */
6 /* This examples shows how to survive to host failure exceptions that occur when an host is turned off.
8 * The actors do not get notified when the host on which they run is turned off: they are just terminated
9 * in this case, and their ``on_exit()`` callback gets executed.
11 * For remote executions on failing hosts however, any blocking operation such as ``exec`` or ``wait`` will
12 * raise an exception that you can catch and react to, as illustrated in this example.
15 #include <simgrid/s4u.hpp>
16 #include "simgrid/kernel/ProfileBuilder.hpp"
18 XBT_LOG_NEW_DEFAULT_CATEGORY(s4u_exec_failure, "Messages specific for this s4u example");
19 namespace sg4 = simgrid::s4u;
21 static void dispatcher(std::vector<sg4::Host*> const& hosts)
23 std::vector<sg4::ExecPtr> pending_execs;
24 for (auto* host: hosts) {
25 XBT_INFO("Initiating asynchronous exec on %s", host->get_cname());
26 // Computing 20 flops on an host which speed is 1f takes 20 seconds (when it does not fail)
27 auto exec = sg4::this_actor::exec_init(20)->set_host(host);
28 pending_execs.push_back(exec);
32 XBT_INFO("---------------------------------");
33 XBT_INFO("Wait on the first exec, which host is turned off at t=10 by the another actor.");
35 pending_execs[0]->wait();
36 xbt_assert("This wait was not supposed to succeed.");
37 } catch (const simgrid::HostFailureException&) {
38 XBT_INFO("Dispatcher has experienced a host failure exception, so it knows that something went wrong.");
41 XBT_INFO("State of each exec:");
42 for (auto const& exec : pending_execs)
43 XBT_INFO(" Exec on %s has state: %s", exec->get_host()->get_cname(), exec->get_state_str());
45 XBT_INFO("---------------------------------");
46 XBT_INFO("Wait on the second exec, which host is turned off at t=12 by the state profile.");
48 pending_execs[1]->wait();
49 xbt_assert("This wait was not supposed to succeed.");
50 } catch (const simgrid::HostFailureException&) {
51 XBT_INFO("Dispatcher has experienced a host failure exception, so it knows that something went wrong.");
53 XBT_INFO("State of each exec:");
54 for (auto const& exec : pending_execs)
55 XBT_INFO(" Exec on %s has state: %s", exec->get_host()->get_cname(), exec->get_state_str());
57 XBT_INFO("---------------------------------");
58 XBT_INFO("Wait on the third exec, which should succeed.");
60 pending_execs[2]->wait();
61 XBT_INFO("No exception occured.");
62 } catch (const simgrid::HostFailureException&) {
63 xbt_assert("This wait was not supposed to fail.");
65 XBT_INFO("State of each exec:");
66 for (auto const& exec : pending_execs)
67 XBT_INFO(" Exec on %s has state: %s", exec->get_host()->get_cname(), exec->get_state_str());
70 static void host_killer(sg4::Host* to_kill)
72 sg4::this_actor::sleep_for(10.0);
73 XBT_INFO("HostKiller turns off the host '%s'.", to_kill->get_cname());
77 int main(int argc, char** argv)
79 sg4::Engine engine(&argc, argv);
81 auto* zone = sg4::create_full_zone("world");
82 std::vector<sg4::Host*> hosts;
83 for (auto name : {"Host1", "Host2", "Host3"}) {
84 auto* host = zone->create_host(name, "1f");
85 hosts.push_back(host);
87 /* Attaching a state profile (ie a list of events changing the on/off state of the resource) to host3.
88 * The syntax of the profile (second parameter) is a list of: "date state\n"
89 * The R"( )" thing is the C++ way of writing multiline strings, including literals \n.
90 * You'd have the same behavior by using "12 0\n20 1\n" instead.
91 * So here, the link is turned off at t=12 and back on at t=20.
92 * The last parameter is the period of that profile, meaning that it loops after 30 seconds.
94 hosts[1]->set_state_profile(simgrid::kernel::profile::ProfileBuilder::from_string("profile name", R"(
101 sg4::Actor::create("Dispatcher", hosts[2], dispatcher, hosts);
102 sg4::Actor::create("HostKiller", hosts[2], host_killer, hosts[0]);