Default MPI stats

#include <gempba/defaults/default_mpi_stats.hpp>  // included automatically via gempba.hpp

The built-in stats implementation for the MPI schedulers. An instance is owned by each worker process during a run. After synchronize_stats(), the center collects one instance from every rank.

---

Fields

std::size_t received_task_count;

Tasks received from other processes during the run.

std::size_t sent_task_count;

Tasks sent to other processes during the run.

std::size_t total_requested_tasks;

Total tasks requested from the load balancer over the course of the run.

std::size_t total_thread_requests;

Total number of times worker threads requested work from the load balancer.

double idle_time;

Cumulative wall-clock time in milliseconds that threads spent idle waiting for work.

double elapsed_time;

Total wall-clock time in milliseconds of the run.

---

Reading the stats

Use default_mpi_stats_visitor to extract these fields after the run:

scheduler->synchronize_stats();

auto stats_vec = scheduler->get_stats_vector();

for (std::size_t rank = 0; rank < stats_vec.size(); ++rank) {
    if (!stats_vec[rank]) continue;

    gempba::default_mpi_stats_visitor v;
    stats_vec[rank]->visit(&v);

    std::cout << "Rank " << rank
              << "  received: "  << v.received_task_count << "\n"
              << "  sent: "      << v.sent_task_count     << "\n"
              << "  idle time: " << v.idle_time           << " ms\n"
              << "  elapsed: "   << v.elapsed_time        << " ms\n";
}

Alternatively, use the lambda overload to access fields directly without a visitor:

stats_vec[rank]->visit([](const std::string& key, std::any&& value) {
    if (key == "elapsed_time")
        std::cout << "elapsed: " << std::any_cast<double>(value) << " ms\n";
});