Add RoundRobinOfSourceRanks strategy

Author: franzpoeschel

include/openPMD/ChunkInfo.hpp

@@ -24,6 +24,7 @@
 
 #include "openPMD/Dataset.hpp" // Offset, Extent
 #include "openPMD/benchmark/mpi/BlockSlicer.hpp"
+#include <memory>
 
 #if openPMD_HAVE_MPI
 #include <mpi.h>
@@ -224,6 +225,16 @@ namespace chunk_assignment
         virtual std::unique_ptr<Strategy> clone() const override;
     };
 
+    struct RoundRobinOfSourceRanks : Strategy
+    {
+        Assignment assign(
+            PartialAssignment,
+            RankMeta const &in,
+            RankMeta const &out) override;
+
+        virtual std::unique_ptr<Strategy> clone() const override;
+    };
+
     /**
      * @brief Strategy that assigns chunks to be read by processes within
      *        the same host that produced the chunk.

src/ChunkInfo.cpp

@@ -24,10 +24,13 @@
 #include "openPMD/auxiliary/Mpi.hpp"
 
 #include <algorithm> // std::sort
+#include <deque>
 #include <iostream>
+#include <iterator>
 #include <list>
 #include <map>
 #include <optional>
+#include <set>
 #include <utility>
 
 #ifdef _WIN32
@@ -195,10 +198,10 @@ namespace chunk_assignment
 
     namespace
     {
-        std::map<std::string, std::list<unsigned int> >
+        std::map<std::string, std::list<unsigned int>>
         ranksPerHost(RankMeta const &rankMeta)
         {
-            std::map<std::string, std::list<unsigned int> > res;
+            std::map<std::string, std::list<unsigned int>> res;
             for (auto const &pair : rankMeta)
             {
                 auto &list = res[pair.second];
@@ -294,6 +297,44 @@ namespace chunk_assignment
         return std::unique_ptr<Strategy>(new RoundRobin);
     }
 
+    Assignment RoundRobinOfSourceRanks::assign(
+        PartialAssignment partialAssignment,
+        RankMeta const &, // ignored parameter
+        RankMeta const &out)
+    {
+        std::map<unsigned int, std::deque<WrittenChunkInfo>>
+            sortSourceChunksBySourceRank;
+        for (auto &chunk : partialAssignment.notAssigned)
+        {
+            auto sourceID = chunk.sourceID;
+            sortSourceChunksBySourceRank[sourceID].push_back(std::move(chunk));
+        }
+        partialAssignment.notAssigned.clear();
+        auto source_it = sortSourceChunksBySourceRank.begin();
+        auto sink_it = out.begin();
+        for (; source_it != sortSourceChunksBySourceRank.end();
+             ++source_it, ++sink_it)
+        {
+            if (sink_it == out.end())
+            {
+                sink_it = out.begin();
+            }
+            auto &chunks_go_here = partialAssignment.assigned[sink_it->first];
+            chunks_go_here.reserve(
+                partialAssignment.assigned.size() + source_it->second.size());
+            for (auto &chunk : source_it->second)
+            {
+                chunks_go_here.push_back(std::move(chunk));
+            }
+        }
+        return partialAssignment.assigned;
+    }
+
+    std::unique_ptr<Strategy> RoundRobinOfSourceRanks::clone() const
+    {
+        return std::unique_ptr<Strategy>(new RoundRobinOfSourceRanks);
+    }
+
     ByHostname::ByHostname(std::unique_ptr<Strategy> withinNode)
         : m_withinNode(std::move(withinNode))
     {}
@@ -332,7 +373,7 @@ namespace chunk_assignment
         // the ranks are the source ranks
 
         // which ranks live on host <string> in the sink?
-        std::map<std::string, std::list<unsigned int> > ranksPerHostSink =
+        std::map<std::string, std::list<unsigned int>> ranksPerHostSink =
             ranksPerHost(out);
         for (auto &chunkGroup : chunkGroups)
         {

src/binding/python/ChunkInfo.cpp

@@ -298,6 +298,8 @@ void init_Chunk(py::module &m)
         }));
 
     py::class_<RoundRobin, Strategy>(m, "RoundRobin").def(py::init<>());
+    py::class_<RoundRobinOfSourceRanks, Strategy>(m, "RoundRobinOfSourceRanks")
+        .def(py::init<>());
 
     py::class_<ByHostname, PartialStrategy>(m, "ByHostname")
         .def(

test/ParallelIOTest.cpp

@@ -1,6 +1,7 @@
 /* Running this test in parallel with MPI requires MPI::Init.
  * To guarantee a correct call to Init, launch the tests manually.
  */
+#include "openPMD/ChunkInfo.hpp"
 #include "openPMD/IO/ADIOS/macros.hpp"
 #include "openPMD/IO/Access.hpp"
 #include "openPMD/auxiliary/Environment.hpp"
@@ -2213,11 +2214,13 @@ void adios2_chunk_distribution()
         series.setRankTable(writingRanksHostnames.at(mpi_rank));
 
         auto E_x = series.iterations[0].meshes["E"]["x"];
-        openPMD::Dataset ds(openPMD::Datatype::INT, {unsigned(mpi_size), 10});
+        openPMD::Dataset ds(
+            openPMD::Datatype::INT, {unsigned(mpi_size * 2), 10});
         E_x.resetDataset(ds);
         std::vector<int> data(10, 0);
         std::iota(data.begin(), data.end(), 0);
-        E_x.storeChunk(data, {unsigned(mpi_rank), 0}, {1, 10});
+        E_x.storeChunk(data, {unsigned(mpi_rank * 2), 0}, {1, 10});
+        E_x.storeChunk(data, {unsigned(mpi_rank * 2 + 1), 0}, {1, 10});
         series.flush();
     }
 
@@ -2276,6 +2279,23 @@ void adios2_chunk_distribution()
             byHostnamePartialAssignment.notAssigned,
             rankMetaIn);
 
+        /*
+         * Same as above, but use RoundRobinOfSourceRanks this time, a strategy
+         * which ensures that each source rank's data is uniquely mapped to one
+         * sink rank. Needed in some domains.
+         */
+        ByHostname byHostname2(std::make_unique<RoundRobinOfSourceRanks>());
+        auto byHostnamePartialAssignment2 =
+            byHostname2.assign(chunkTable, rankMetaIn, readingRanksHostnames);
+        printAssignment(
+            "HOSTNAME2, ASSIGNED",
+            byHostnamePartialAssignment2.assigned,
+            readingRanksHostnames);
+        printChunktable(
+            "HOSTNAME2, LEFTOVER",
+            byHostnamePartialAssignment2.notAssigned,
+            rankMetaIn);
+
         /*
          * Assign chunks by hostnames, once more.
          * This time, apply a secondary distribution strategy to assign