Ejecuta NCCL en clústeres personalizados de GKE que usan A4 o A3 Ultra

En esta página, se describe cómo ejecutar pruebas de NCCL/gIB en clústeres aprovisionados que usan GPUDirect RDMA. Describe las pruebas para las siguientes situaciones:

Prueba en dos nodos

Ejecuta la prueba de dos nodos:

A4

  1. Para implementar una carga de trabajo de prueba de NCCL de dos Pods de prueba que se ejecutan en dos nodos A4, aplica uno de los siguientes manifiestos:

    • Para un clúster de Autopilot:

      kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/refs/heads/master/gpudirect-rdma/nccl-test-a4-autopilot.yaml
      
    • Para un clúster de Standard:

      kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/refs/heads/master/gpudirect-rdma/nccl-test-a4.yaml
      
  2. Verifica si los Pods están programados y se ejecutan en algunos nodos:

    kubectl get pods nccl-test-host-1 nccl-test-host-2
    

    Si los dos Pods tienen el estado Running, puedes continuar con el siguiente paso. En el caso de los nodos aprovisionados por flex-start, es posible que pasen algunos minutos antes de que se creen los nodos y se programen los Pods en ellos.

  3. Activa una prueba de recopilación completa de NCCL para los nodos:

    kubectl exec nccl-test-host-1 -it -- /usr/local/gib/scripts/run_nccl_tests.sh -t all_gather -b 1K -e 8G nccl-host-1 nccl-host-2
    

    El resultado debería ser similar al siguiente ejemplo:

    #       size         count      type   redop    root     time   algbw   busbw #wrong     time   algbw   busbw #wrong
    #        (B)    (elements)                               (us)  (GB/s)  (GB/s)            (us)  (GB/s)  (GB/s)
            1024            16     float    none      -1    48.17    0.02    0.02      0    47.21    0.02    0.02      0
            2048            32     float    none      -1    47.23    0.04    0.04      0    47.17    0.04    0.04      0
            4096            64     float    none      -1    47.43    0.09    0.08      0    47.48    0.09    0.08      0
            8192           128     float    none      -1    47.93    0.17    0.16      0    47.98    0.17    0.16      0
           16384           256     float    none      -1    48.90    0.34    0.31      0    48.75    0.34    0.32      0
           32768           512     float    none      -1    50.10    0.65    0.61      0    49.59    0.66    0.62      0
           65536          1024     float    none      -1    51.70    1.27    1.19      0    51.66    1.27    1.19      0
          131072          2048     float    none      -1    52.23    2.51    2.35      0    55.60    2.36    2.21      0
          262144          4096     float    none      -1    53.89    4.86    4.56      0    53.39    4.91    4.60      0
          524288          8192     float    none      -1    56.80    9.23    8.65      0    57.66    9.09    8.52      0
         1048576         16384     float    none      -1    87.85   11.94   11.19      0    88.47   11.85   11.11      0
         2097152         32768     float    none      -1    92.52   22.67   21.25      0    93.22   22.50   21.09      0
         4194304         65536     float    none      -1    97.41   43.06   40.37      0    96.15   43.62   40.90      0
         8388608        131072     float    none      -1    110.0   76.27   71.51      0    110.9   75.66   70.93      0
        16777216        262144     float    none      -1    141.3  118.77  111.35      0    140.7  119.27  111.81      0
        33554432        524288     float    none      -1    203.2  165.14  154.82      0    202.3  165.90  155.53      0
        67108864       1048576     float    none      -1    303.3  221.25  207.42      0    301.9  222.27  208.38      0
       134217728       2097152     float    none      -1    513.2  261.56  245.21      0    509.3  263.56  247.08      0
       268435456       4194304     float    none      -1    842.4  318.64  298.72      0    832.3  322.54  302.38      0
       536870912       8388608     float    none      -1   1511.8  355.12  332.92      0   1502.5  357.31  334.98      0
      1073741824      16777216     float    none      -1   2976.7  360.72  338.17      0   2923.2  367.32  344.36      0
      2147483648      33554432     float    none      -1   5888.9  364.66  341.87      0   5766.2  372.43  349.15      0
      4294967296      67108864     float    none      -1    11722  366.39  343.49      0    11457  374.88  351.45      0
      8589934592     134217728     float    none      -1    23379  367.43  344.46      0    22818  376.45  352.92      0
    # Out of bounds values : 0 OK
    # Avg bus bandwidth    : 120.845
    

A3 Ultra

  1. Para implementar una carga de trabajo de prueba de NCCL de dos Pods de prueba que se ejecutan en dos nodos A3 Ultra, aplica uno de los siguientes manifiestos:

    • Para un clúster de Autopilot:

      kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/refs/heads/master/gpudirect-rdma/nccl-test-autopilot.yaml
      
    • Para un clúster de Standard:

      kubectl apply -f https://raw.githubusercontent.com/GoogleCloudPlatform/container-engine-accelerators/refs/heads/master/gpudirect-rdma/nccl-test.yaml
      
  2. Verifica si los Pods están programados y se ejecutan en algunos nodos:

    kubectl get pods nccl-test-host-1 nccl-test-host-2
    

    Si los dos Pods tienen el estado Running, puedes continuar con el siguiente paso. En el caso de los nodos aprovisionados por flex-start, es posible que pasen unos minutos antes de que se creen los nodos y se programen los Pods en ellos.

  3. Activa una prueba de recopilación completa de NCCL para los nodos:

    kubectl exec nccl-test-host-1 -it -- /usr/local/gib/scripts/run_nccl_tests.sh -t all_gather -b 1K -e 8G nccl-host-1 nccl-host-2
    

    El resultado debería ser similar al siguiente ejemplo:

    #       size         count      type   redop    root     time   algbw   busbw #wrong     time   algbw   busbw #wrong
    #        (B)    (elements)                               (us)  (GB/s)  (GB/s)            (us)  (GB/s)  (GB/s)
            1024            16     float    none      -1    56.00    0.02    0.02      0    55.59    0.02    0.02      0
            2048            32     float    none      -1    55.79    0.04    0.03      0    55.57    0.04    0.03      0
            4096            64     float    none      -1    56.29    0.07    0.07      0    57.35    0.07    0.07      0
            8192           128     float    none      -1    56.44    0.15    0.14      0    56.32    0.15    0.14      0
           16384           256     float    none      -1    57.57    0.28    0.27      0    57.60    0.28    0.27      0
           32768           512     float    none      -1    57.92    0.57    0.53      0    59.35    0.55    0.52      0
           65536          1024     float    none      -1    59.92    1.09    1.03      0    60.15    1.09    1.02      0
          131072          2048     float    none      -1    59.21    2.21    2.08      0    61.82    2.12    1.99      0
          262144          4096     float    none      -1    63.58    4.12    3.87      0    63.34    4.14    3.88      0
          524288          8192     float    none      -1    64.89    8.08    7.57      0    65.09    8.06    7.55      0
         1048576         16384     float    none      -1    80.90   12.96   12.15      0    77.49   13.53   12.69      0
         2097152         32768     float    none      -1    80.22   26.14   24.51      0    79.88   26.25   24.61      0
         4194304         65536     float    none      -1    82.86   50.62   47.45      0    82.47   50.86   47.68      0
         8388608        131072     float    none      -1    95.83   87.53   82.06      0    93.27   89.94   84.32      0
        16777216        262144     float    none      -1    122.8  136.58  128.04      0    121.7  137.86  129.24      0
        33554432        524288     float    none      -1    180.6  185.75  174.14      0    179.2  187.19  175.49      0
        67108864       1048576     float    none      -1    279.7  239.90  224.90      0    277.0  242.26  227.12      0
       134217728       2097152     float    none      -1    507.5  264.46  247.93      0    485.1  276.66  259.37      0
       268435456       4194304     float    none      -1    866.3  309.88  290.51      0    864.0  310.70  291.28      0
       536870912       8388608     float    none      -1   1576.1  340.62  319.33      0   1558.2  344.54  323.01      0
      1073741824      16777216     float    none      -1   3096.6  346.75  325.08      0   3047.5  352.33  330.31      0
      2147483648      33554432     float    none      -1   6148.0  349.30  327.47      0   6034.3  355.88  333.64      0
      4294967296      67108864     float    none      -1    12226  351.29  329.33      0    12000  357.92  335.55      0
      8589934592     134217728     float    none      -1    24391  352.17  330.16      0    23920  359.11  336.67      0
    # Out of bounds values : 0 OK
    # Avg bus bandwidth    : 120.94
    

Prueba con la programación consciente de la topología (TAS)

Si tienes más de dos nodos, te recomendamos que uses la siguiente prueba, que utiliza TAS. Sigue los pasos que se indican en las próximas secciones para preparar y ejecutar la prueba en tu clúster.

Configura tu clúster con Jobset y el complemento de TAS

  1. Instala JobSet.

  2. Instala el complemento de TAS:

    1. Clona el repositorio de Git container-engine-accelerators:

      cd ~
      git clone https://github.com/GoogleCloudPlatform/container-engine-accelerators.git
      
    2. Aplica el complemento de TAS:

      cd container-engine-accelerators/gke-topology-scheduler
      kubectl create configmap topology-scheduler-scripts --namespace kube-system --from-file=schedule-daemon.py=schedule-daemon.py --from-file=label-nodes-daemon.py=label-nodes-daemon.py
      kubectl apply -f service-account.yaml
      kubectl apply -f schedule-daemon.yaml
      kubectl apply -f label-nodes-daemon.yaml
      

Implementa una carga de trabajo de prueba de NCCL con TAS

A4

  1. Crea el siguiente manifiesto nccl-jobset-test.yaml:

    apiVersion: jobset.x-k8s.io/v1alpha2
    kind: JobSet
    metadata:
      # The name `nccl-ag` is used for an NCCL all-gather test.
      name: nccl-ag
    spec:
      ttlSecondsAfterFinished: 1200
      suspend: False
      network:
        enableDNSHostnames: true
      replicatedJobs:
        - name: worker
          template:
            spec:
              parallelism: NUM_NODES
              completions: NUM_NODES
              template:
                metadata:
                  annotations:
                    networking.gke.io/default-interface: 'eth0'
                    networking.gke.io/interfaces: |
                      [
                        {"interfaceName":"eth0","network":"default"},
                        {"interfaceName":"eth2","network":"rdma-0"},
                        {"interfaceName":"eth3","network":"rdma-1"},
                        {"interfaceName":"eth4","network":"rdma-2"},
                        {"interfaceName":"eth5","network":"rdma-3"},
                        {"interfaceName":"eth6","network":"rdma-4"},
                        {"interfaceName":"eth7","network":"rdma-5"},
                        {"interfaceName":"eth8","network":"rdma-6"},
                        {"interfaceName":"eth9","network":"rdma-7"}
                      ]
                spec:
                  activeDeadlineSeconds: 3600
                  restartPolicy: Never
                  nodeSelector:
                    cloud.google.com/gke-accelerator: nvidia-b200
                  tolerations:
                  - key: cloud.google.com/gke-queued
                    effect: NoSchedule
                    value: "true"
                  - key: "nvidia.com/gpu"
                    operator: "Exists"
                    effect: "NoSchedule"
                  setHostnameAsFQDN: true
                  volumes:
                  - name: gib
                    hostPath:
                      path: /home/kubernetes/bin/gib
                  - name: nvidia
                    hostPath:
                      path: /home/kubernetes/bin/nvidia
                  - name: lib64
                    hostPath:
                      path: /lib64
                  - name: shared-memory
                    emptyDir:
                      medium: "Memory"
                      sizeLimit: 250Gi
                  schedulingGates:
                  - name: "gke.io/topology-aware-auto-nccl-test"
                  containers:
                  - name: nccl-test
                    stdin: true
                    tty: true
                    image: us-docker.pkg.dev/gce-ai-infra/gpudirect-gib/nccl-plugin-gib-diagnostic:v1.0.6
                    env:
                    - name: MY_NODE_NAME
                      valueFrom:
                        fieldRef:
                          fieldPath: spec.nodeName
                    - name: OMPI_ALLOW_RUN_AS_ROOT
                      value: "1"
                    - name: OMPI_ALLOW_RUN_AS_ROOT_CONFIRM
                      value: "1"
                    - name: N_NODES
                      value: "NUM_NODES"
                    - name: LD_LIBRARY_PATH
                      value: /usr/local/nvidia/lib64
                    command:
                    - bash
                    - -c
                    - |
                      set -x
                      echo "Starting workload container on ${MY_NODE_NAME} for $N_NODES benchmark"
                      # Install ping
                      apt update -y
                      apt install -y iputils-ping
    
                      # Start sshd
                      /scripts/container_entry.sh daemon &
    
                      # Get helper variables to form all hostnames
                      export POSTFIX=$(hostname | cut -d . -f 2-)
                      export WORKERS_BASENAME=$(hostname | cut -d . -f 1 | rev | cut -d - -f 2- | rev )
                      export NODE_RANK=$JOB_COMPLETION_INDEX
    
                      # For every worker, wait till online and add to hostfile
                      for i in `seq 0 $(($N_NODES-1))`; do
                        OTHER=${WORKERS_BASENAME}-${i}.${POSTFIX}
                        until ssh -p 222 -o StrictHostKeyChecking=no $OTHER hostname; do
                          echo Waiting for ${OTHER}...
                          sleep 10
                        done
                        echo ${OTHER} port=222 slots=8 | tee -a /tmp/hostfile;
                      done
    
                      cat /tmp/hostfile
    
                      # Launch from head node
                      if [[ "${NODE_RANK}" -eq "0" ]]; then
    
                          # World Level = 0x0, Rail Aligned = 0x7
                          export NCCL_TESTS_SPLIT_MASK="0x0";
    
                          # Force use of libnccl-gib
                          export NCCL_NET=gIB
    
                          # Set all the correct libnccl-gib environment variables
                          source /usr/local/gib/scripts/set_nccl_env.sh
    
                          # Get all relevant NCCL / env vars to pass to all workers
                          ENV_VARS=$(echo ${!NCCL*} ${!OMPI*} LD_LIBRARY_PATH PATH | sed 's/ / -x /g')
    
                          mpirun --hostfile /tmp/hostfile \
                            -x $ENV_VARS  \
                            -mca plm_rsh_no_tree_spawn 1 \
                            --mca mtl ^ofi \
                            --mca orte_keep_fqdn_hostnames 1 \
                            --mca btl self,tcp \
                            --mca btl_tcp_if_include eth0 \
                            --bind-to none \
                            --mca plm_rsh_agent "ssh -q -o LogLevel=ERROR -o StrictHostKeyChecking=no -p 222" \
                            /third_party/nccl-tests/build/all_gather_perf -b 1K -e 8G -f 2 -g 1 -w 5 --iters 100 -c 1
    
                      else
                          while ping -c 1 ${WORKERS_BASENAME}-0.${POSTFIX}; do
                          sleep 5
                      done
                      fi
    
                      exit 0
                    volumeMounts:
                    - name: nvidia
                      mountPath: /usr/local/nvidia
                    - name: gib
                      mountPath: /usr/local/gib
                    - name: shared-memory
                      mountPath: /dev/shm
                    resources:
                      limits:
                        nvidia.com/gpu: 8
                      requests:
                        nvidia.com/gpu: 8
                  restartPolicy: Never
    

    Reemplaza NUM_NODES por la cantidad de nodos del grupo de nodos.

    Asegúrate de comprender lo siguiente sobre este manifiesto:

    • El JobSet es un Service sin interfaz gráfica con el mismo nombre que el nombre del JobSet; en este caso, nccl-ag.
    • La puerta de programación gke.io/topology-aware-auto-nccl-test se usa para verificar que los Pods estén programados para la colocación.
    • Los campos parallelism y completions se establecen en la cantidad de nodos que deseas usar para ejecutar la prueba de NCCL.
  2. Aplica el manifiesto

    kubectl apply -f nccl-jobset-test.yaml
    
  3. Confirma que se admita la carga de trabajo:

    kubectl get jobsets
    

    El resultado es similar a lo siguiente:

    NAME            RESTARTS   COMPLETED   AGE
    nccl-ag                                3s
    
  4. Confirma que la carga de trabajo esté en el estado Completed:

    kubectl get pods
    

    El resultado es similar a lo siguiente:

    NAME                       READY   STATUS      RESTARTS   AGE
    nccl-ag-worker-0-0-n9s6j   0/1     Completed   0          9m34s
    nccl-ag-worker-0-1-rsf7r   0/1     Completed   0          9m34s
    ...
    
  5. Los registros del Pod con el patrón nccl-ag-worker-0-0-.* contienen los resultados de la prueba.

    Recupera los registros de este Pod:

      kubectl logs $(kubectl get pods -o go-template='{{range .items}}{{.metadata.name}}{{"\n"}}{{end}}' | grep nccl-ag-worker-0-0)
    

    El resultado debería ser similar al siguiente ejemplo:

    #       size         count      type   redop    root     time   algbw   busbw #wrong     time   algbw   busbw #wrong
    #        (B)    (elements)                               (us)  (GB/s)  (GB/s)            (us) ∂ç (GB/s)  (GB/s)
            1024            16     float    none      -1    54.07    0.02    0.02      0    55.80    0.02    0.02      0
            2048            32     float    none      -1    55.46    0.04    0.03      0    55.31    0.04    0.03      0
            4096            64     float    none      -1    55.59    0.07    0.07      0    55.38    0.07    0.07      0
            8192           128     float    none      -1    56.05    0.15    0.14      0    55.92    0.15    0.14      0
           16384           256     float    none      -1    57.08    0.29    0.27      0    57.75    0.28    0.27      0
           32768           512     float    none      -1    57.49    0.57    0.53      0    57.22    0.57    0.54      0
           65536          1024     float    none      -1    59.20    1.11    1.04      0    59.20    1.11    1.04      0
          131072          2048     float    none      -1    59.58    2.20    2.06      0    63.57    2.06    1.93      0
          262144          4096     float    none      -1    63.87    4.10    3.85      0    63.61    4.12    3.86      0
          524288          8192     float    none      -1    64.83    8.09    7.58      0    64.40    8.14    7.63      0
         1048576         16384     float    none      -1    79.74   13.15   12.33      0    76.66   13.68   12.82      0
         2097152         32768     float    none      -1    78.41   26.74   25.07      0    79.05   26.53   24.87      0
         4194304         65536     float    none      -1    83.21   50.41   47.26      0    81.25   51.62   48.39      0
         8388608        131072     float    none      -1    94.35   88.91   83.35      0    99.07   84.68   79.38      0
        16777216        262144     float    none      -1    122.9  136.55  128.02      0    121.7  137.83  129.21      0
        33554432        524288     float    none      -1    184.2  182.19  170.80      0    178.1  188.38  176.60      0
        67108864       1048576     float    none      -1    294.7  227.75  213.51      0    277.7  241.62  226.52      0
       134217728       2097152     float    none      -1    495.4  270.94  254.00      0    488.8  274.60  257.43      0
       268435456       4194304     float    none      -1    877.5  305.92  286.80      0    861.3  311.65  292.17      0
       536870912       8388608     float    none      -1   1589.8  337.71  316.60      0   1576.2  340.61  319.33      0
      1073741824      16777216     float    none      -1   3105.7  345.74  324.13      0   3069.2  349.85  327.98      0
      2147483648      33554432     float    none      -1   6161.7  348.52  326.74      0   6070.7  353.75  331.64      0
      4294967296      67108864     float    none      -1    12305  349.03  327.22      0    12053  356.35  334.08      0
      8589934592     134217728     float    none      -1    24489  350.77  328.85      0    23991  358.05  335.67      0
    # Out of bounds values : 0 OK
    # Avg bus bandwidth    : 120.248
    

A3 Ultra

  1. Crea el siguiente manifiesto nccl-jobset-test.yaml:

    apiVersion: jobset.x-k8s.io/v1alpha2
    kind: JobSet
    metadata:
      # The name `nccl-ag` is used for an NCCL all-gather test.
      name: nccl-ag
    spec:
      ttlSecondsAfterFinished: 1200
      suspend: False
      network:
        enableDNSHostnames: true
      replicatedJobs:
        - name: worker
          template:
            spec:
              parallelism: NUM_NODES
              completions: NUM_NODES
              template:
                metadata:
                  annotations:
                    networking.gke.io/default-interface: 'eth0'
                    networking.gke.io/interfaces: |
                      [
                        {"interfaceName":"eth0","network":"default"},
                        {"interfaceName":"eth2","network":"rdma-0"},
                        {"interfaceName":"eth3","network":"rdma-1"},
                        {"interfaceName":"eth4","network":"rdma-2"},
                        {"interfaceName":"eth5","network":"rdma-3"},
                        {"interfaceName":"eth6","network":"rdma-4"},
                        {"interfaceName":"eth7","network":"rdma-5"},
                        {"interfaceName":"eth8","network":"rdma-6"},
                        {"interfaceName":"eth9","network":"rdma-7"}
                      ]
                spec:
                  activeDeadlineSeconds: 3600
                  restartPolicy: Never
                  nodeSelector:
                    cloud.google.com/gke-accelerator: nvidia-h200-141gb
                  tolerations:
                  - key: cloud.google.com/gke-queued
                    effect: NoSchedule
                    value: "true"
                  - key: "nvidia.com/gpu"
                    operator: "Exists"
                    effect: "NoSchedule"
                  setHostnameAsFQDN: true
                  volumes:
                  - name: gib
                    hostPath:
                      path: /home/kubernetes/bin/gib
                  - name: nvidia
                    hostPath:
                      path: /home/kubernetes/bin/nvidia
                  - name: lib64
                    hostPath:
                      path: /lib64
                  - name: shared-memory
                    emptyDir:
                      medium: "Memory"
                      sizeLimit: 250Gi
                  schedulingGates:
                  - name: "gke.io/topology-aware-auto-nccl-test"
                  containers:
                  - name: nccl-test
                    stdin: true
                    tty: true
                    image: us-docker.pkg.dev/gce-ai-infra/gpudirect-gib/nccl-plugin-gib-diagnostic:v1.0.6
                    securityContext:
                      privileged: true
                    env:
                    - name: MY_NODE_NAME
                      valueFrom:
                        fieldRef:
                          fieldPath: spec.nodeName
                    - name: OMPI_ALLOW_RUN_AS_ROOT
                      value: "1"
                    - name: OMPI_ALLOW_RUN_AS_ROOT_CONFIRM
                      value: "1"
                    - name: N_NODES
                      value: "NUM_NODES"
                    - name: LD_LIBRARY_PATH
                      value: /usr/local/nvidia/lib64
                    command:
                    - bash
                    - -c
                    - |
                      set -x
                      echo "Starting workload container on ${MY_NODE_NAME} for $N_NODES benchmark"
                      # Install ping
                      apt update -y
                      apt install -y iputils-ping
    
                      # Start sshd
                      /scripts/container_entry.sh daemon &
    
                      # Get helper variables to form all hostnames
                      export POSTFIX=$(hostname | cut -d . -f 2-)
                      export WORKERS_BASENAME=$(hostname | cut -d . -f 1 | rev | cut -d - -f 2- | rev )
                      export NODE_RANK=$JOB_COMPLETION_INDEX
    
                      # For every worker, wait till online and add to hostfile
                      for i in `seq 0 $(($N_NODES-1))`; do
                        OTHER=${WORKERS_BASENAME}-${i}.${POSTFIX}
                        until ssh -p 222 -o StrictHostKeyChecking=no $OTHER hostname; do
                          echo Waiting for ${OTHER}...
                          sleep 10
                        done
                        echo ${OTHER} port=222 slots=8 | tee -a /tmp/hostfile;
                      done
    
                      cat /tmp/hostfile
    
                      # Launch from head node
                      if [[ "${NODE_RANK}" -eq "0" ]]; then
    
                          # World Level = 0x0, Rail Aligned = 0x7
                          export NCCL_TESTS_SPLIT_MASK="0x0";
    
                          # Force use of libnccl-gib
                          export NCCL_NET=gIB
    
                          # Set all the correct libnccl-gib environment variables
                          source /usr/local/gib/scripts/set_nccl_env.sh
    
                          # Get all relevant NCCL / env vars to pass to all workers
                          ENV_VARS=$(echo ${!NCCL*} ${!OMPI*} LD_LIBRARY_PATH PATH | sed 's/ / -x /g')
    
                          mpirun --hostfile /tmp/hostfile \
                            -x $ENV_VARS  \
                            -mca plm_rsh_no_tree_spawn 1 \
                            --mca orte_keep_fqdn_hostnames 1 \
                            --mca btl self,tcp \
                            --mca btl_tcp_if_include eth0 \
                            --bind-to none \
                            --mca plm_rsh_agent "ssh -q -o LogLevel=ERROR -o StrictHostKeyChecking=no -p 222" \
                            /third_party/nccl-tests/build/all_gather_perf -b 1K -e 8G -f 2 -g 1 -w 5 --iters 100 -c 1
    
                      else
                          while ping -c 1 ${WORKERS_BASENAME}-0.${POSTFIX}; do
                          sleep 5
                      done
                      fi
    
                      exit 0
                    volumeMounts:
                    - name: nvidia
                      mountPath: /usr/local/nvidia
                    - name: gib
                      mountPath: /usr/local/gib
                    - name: shared-memory
                      mountPath: /dev/shm
                    resources:
                      limits:
                        nvidia.com/gpu: 8
                      requests:
                        nvidia.com/gpu: 8
                  restartPolicy: Never
    

    Reemplaza NUM_NODES por la cantidad de nodos del grupo de nodos.

    Asegúrate de comprender lo siguiente sobre este manifiesto:

    • El JobSet es un Service sin interfaz gráfica con el mismo nombre que el nombre del JobSet; en este caso, nccl-ag.
    • La puerta de programación gke.io/topology-aware-auto-nccl-test se usa para verificar que los Pods estén programados para la colocación.
    • Los campos parallelism y completions se configuran en la cantidad de nodos que deseas usar para ejecutar la prueba de NCCL.
  2. Aplica el manifiesto

    kubectl apply -f nccl-jobset-test.yaml
    
  3. Confirma que se admita la carga de trabajo:

    kubectl get jobsets
    

    El resultado es similar a lo siguiente:

    NAME            RESTARTS   COMPLETED   AGE
    nccl-ag                                3s
    
  4. Confirma que la carga de trabajo esté en el estado Completed:

    kubectl get pods
    

    El resultado es similar a lo siguiente:

    NAME                       READY   STATUS      RESTARTS   AGE
    nccl-ag-worker-0-0-n9s6j   0/1     Completed   0          9m34s
    nccl-ag-worker-0-1-rsf7r   0/1     Completed   0          9m34s
    ...
    
  5. Los registros del Pod con el patrón nccl-ag-worker-0-0-.* contienen los resultados de la prueba.

    Recupera los registros de este Pod:

      kubectl logs $(kubectl get pods -o go-template='{{range .items}}{{.metadata.name}}{{"\n"}}{{end}}' | grep nccl-ag-worker-0-0)
    

    El resultado debería ser similar al siguiente ejemplo:

      #       size         count      type   redop    root     time   algbw   busbw #wrong     time   algbw   busbw #wrong
      #        (B)    (elements)                               (us)  (GB/s)  (GB/s)            (us) ∂ç (GB/s)  (GB/s)
              1024            16     float    none      -1    54.07    0.02    0.02      0    55.80    0.02    0.02      0
              2048            32     float    none      -1    55.46    0.04    0.03      0    55.31    0.04    0.03      0
              4096            64     float    none      -1    55.59    0.07    0.07      0    55.38    0.07    0.07      0
              8192           128     float    none      -1    56.05    0.15    0.14      0    55.92    0.15    0.14      0
             16384           256     float    none      -1    57.08    0.29    0.27      0    57.75    0.28    0.27      0
             32768           512     float    none      -1    57.49    0.57    0.53      0    57.22    0.57    0.54      0
             65536          1024     float    none      -1    59.20    1.11    1.04      0    59.20    1.11    1.04      0
            131072          2048     float    none      -1    59.58    2.20    2.06      0    63.57    2.06    1.93      0
            262144          4096     float    none      -1    63.87    4.10    3.85      0    63.61    4.12    3.86      0
            524288          8192     float    none      -1    64.83    8.09    7.58      0    64.40    8.14    7.63      0
           1048576         16384     float    none      -1    79.74   13.15   12.33      0    76.66   13.68   12.82      0
           2097152         32768     float    none      -1    78.41   26.74   25.07      0    79.05   26.53   24.87      0
           4194304         65536     float    none      -1    83.21   50.41   47.26      0    81.25   51.62   48.39      0
           8388608        131072     float    none      -1    94.35   88.91   83.35      0    99.07   84.68   79.38      0
          16777216        262144     float    none      -1    122.9  136.55  128.02      0    121.7  137.83  129.21      0
          33554432        524288     float    none      -1    184.2  182.19  170.80      0    178.1  188.38  176.60      0
          67108864       1048576     float    none      -1    294.7  227.75  213.51      0    277.7  241.62  226.52      0
         134217728       2097152     float    none      -1    495.4  270.94  254.00      0    488.8  274.60  257.43      0
         268435456       4194304     float    none      -1    877.5  305.92  286.80      0    861.3  311.65  292.17      0
         536870912       8388608     float    none      -1   1589.8  337.71  316.60      0   1576.2  340.61  319.33      0
        1073741824      16777216     float    none      -1   3105.7  345.74  324.13      0   3069.2  349.85  327.98      0
        2147483648      33554432     float    none      -1   6161.7  348.52  326.74      0   6070.7  353.75  331.64      0
        4294967296      67108864     float    none      -1    12305  349.03  327.22      0    12053  356.35  334.08      0
        8589934592     134217728     float    none      -1    24489  350.77  328.85      0    23991  358.05  335.67      0
      # Out of bounds values : 0 OK
      # Avg bus bandwidth    : 120.248
    ```
    

¿Qué sigue?