Add fluid flow wan simulation model

.github/workflows/build.yml

@@ -84,6 +84,7 @@ jobs:
       # ROSS install prefix; find_package(ROSS) honors ROSS_ROOT, so the CODES
       # presets need no hardcoded path. Matches the install prefix below.
       ROSS_ROOT: ${{ github.workspace }}/ross-install
+      UCX_TLS: "self,tcp"
     steps:
       - name: Checkout CODES
         uses: actions/checkout@v4
@@ -113,6 +114,17 @@ jobs:
             sudo apt-get install -y clang
           fi
 
+          echo "MPI diagnostics:"
+          which mpiexec || true
+          which mpiexec.mpich || true
+          which mpirun || true
+          which mpirun.mpich || true
+          which mpicc || true
+          mpiexec --version || true
+          mpiexec.mpich --version || true
+          mpicc -show || true
+          dpkg -l | grep -E 'mpich|ucx|openmpi' || true
+
       - name: Install system dependencies (macOS)
         if: runner.os == 'macOS'
         run: |
@@ -282,6 +294,7 @@ jobs:
     env:
       # ROSS install prefix; find_package(ROSS) honors ROSS_ROOT (CODES presets).
       ROSS_ROOT: ${{ github.workspace }}/ross-install
+      UCX_TLS: "self,tcp"
       # detect_leaks=0: corruption detection first; leaks are the noisiest part
       # (MPICH internals + CODES's known unfreed globals) and come later with a
       # suppression file. abort_on_error=1: guarantee a nonzero exit so ctest
@@ -379,6 +392,7 @@ jobs:
     env:
       # ROSS install prefix; find_package(ROSS) honors ROSS_ROOT (CODES presets).
       ROSS_ROOT: ${{ github.workspace }}/ross-install
+      UCX_TLS: "self,tcp"
     steps:
       - name: Checkout CODES
         uses: actions/checkout@v4

doc/example/CMakeLists.txt

@@ -14,6 +14,7 @@ configure_file(tutorial-ping-pong-surrogate.conf.in tutorial-ping-pong-surrogate
 configure_file(kb.dfdally-72-zeromq-director.conf.in kb.dfdally-72-zeromq-director.template.conf.in @ONLY)
 configure_file(kb.dfdally-72-event-time-director.conf.in kb.dfdally-72-event-time-director.template.conf.in @ONLY)
 configure_file(kb.dfdally-72-milc-small.workload.conf.in kb.dfdally-72-milc-small.workload.template.conf.in @ONLY)
+configure_file(fluid-flow-wan.conf.in fluid-flow-wan.template.conf.in @ONLY)
 configure_file(kb.dfdally-72-zeromq-director.conf.in kb.dfdally-72-zeromq-director.template.conf.in @ONLY)
 
 set(single_quote "'")
@@ -42,6 +43,8 @@ configure_file(tutorial-ping-pong-surrogate.conf.in tutorial-ping-pong-surrogate
 configure_file(kb.dfdally-72-zeromq-director.conf.in kb.dfdally-72-zeromq-director.conf)
 configure_file(kb.dfdally-72-event-time-director.conf.in kb.dfdally-72-event-time-director.conf)
 configure_file(kb.dfdally-72-milc-small.workload.conf.in kb.dfdally-72-milc-small.workload.conf)
+configure_file(fluid-flow-wan.conf.in fluid-flow-wan.conf)
 configure_file(kb.dfdally-72-milc-small.json kb.dfdally-72-milc-small.json COPYONLY)
 configure_file(kb.dfdally-72-milc-small.alloc.conf kb.dfdally-72-milc-small.alloc.conf COPYONLY)
+configure_file(fluid-flow-wan-topology.yaml fluid-flow-wan-topology.yaml COPYONLY)
 configure_file(kb.dfdally-72-zeromq-director.conf.in kb.dfdally-72-zeromq-director.conf)

doc/example/fluid-flow-wan-topology.yaml

@@ -0,0 +1,24 @@
+topology:
+  switches:
+    A:
+      terminals: 2
+      terminal_bandwidth: "20 Mbps"
+      switch_buffer: "1000 Mb"
+      connections:
+        B: "15 Mbps"
+
+    B:
+      terminals: 2
+      terminal_bandwidth: "25 Mbps"
+      switch_buffer: "1000 Mb"
+      connections:
+        A: "15 Mbps"
+        C: "10 Mbps"
+
+    C:
+      terminals: 2
+      terminal_bandwidth: "20 Mbps"
+      switch_buffer: "1000 Mb"
+      connections:
+        A: "10 Mbps"
+        B: "15 Mbps"

doc/example/fluid-flow-wan.conf.in

@@ -0,0 +1,45 @@
+# Interval-fluid switch/terminal pure-PDES example.
+# Run from the build directory, for example:
+#   mpirun -np 1 ./src/model-net-fluid-flow-wan --synch=1 -- doc/example/fluid-flow-wan.conf
+
+LPGROUPS
+{
+    FLUID_FLOW_WAN_GRP
+    {
+        repetitions="1";
+        fluid-flow-wan-switch-lp="3";
+        fluid-flow-wan-terminal-lp="6";
+    }
+}
+
+PARAMS
+{
+    message_size="32768";
+    pe_mem_factor="1024";
+}
+
+FLUID_FLOW_WAN
+{
+    topology_yaml_file="fluid-flow-wan-topology.yaml";
+
+    interval_seconds="1";
+    num_send_intervals="20";
+    num_drain_intervals="20";
+    rng_seed="12345";
+
+    terminal_send_every_n_intervals="1";
+    terminal_send_probability="1.0";
+    terminal_min_send_mbit="0";
+    terminal_max_send_fraction_of_link_capacity="1.0";
+
+    debug_prints="0";
+    # This can be either round_robin or fifo.
+    switch_scheduler="round_robin";
+    round_robin_max_entries_per_egress="8";
+    round_robin_quantum_mbit="0";
+
+    terminal_log_path="../../zmqml_artifacts/fluid-flow-wan/terminal-events.csv";
+    switch_log_path="../../zmqml_artifacts/fluid-flow-wan/switch-events.csv";
+    flowlet_log_path="../../zmqml_artifacts/fluid-flow-wan/flowlet-events.csv";
+    switch_training_log_path="../../zmqml_artifacts/fluid-flow-wan/switch-training.csv";
+}

src/CMakeLists.txt

@@ -206,6 +206,7 @@ add_executable(model-net-synthetic network-workloads/model-net-synthetic.c)
 add_executable(model-net-synthetic-slimfly network-workloads/model-net-synthetic-slimfly.c)
 add_executable(model-net-synthetic-fattree network-workloads/model-net-synthetic-fattree.c)
 add_executable(model-net-synthetic-dragonfly-all network-workloads/model-net-synthetic-dragonfly-all.c)
+add_executable(model-net-fluid-flow-wan network-workloads/model-net-fluid-flow-wan.cxx)
 
 set(CODES_TARGETS
     topology-test
@@ -214,6 +215,7 @@ set(CODES_TARGETS
     model-net-synthetic-slimfly
     model-net-synthetic-fattree
     model-net-synthetic-dragonfly-all
+    model-net-fluid-flow-wan
 )
 
 if(USE_DUMPI)