MPI Operator for Kubernetes
On this Page
MPI Operator for Kubernetes¶
Intel® Gaudi® uses the standard MPI Operator from Kubeflow that allows
running MPI allreduce
style workloads in Kubernetes and leveraging Gaudi
accelerators. In combination with Intel Gaudi hardware and software, it
enables large scale distributed training with simple Kubernetes job
distribution model.
Prerequisites¶
Kubernetes version listed in the Support Matrix.
Intel Gaudi software drivers loaded on the system. For more details, refer to Installation Guide.
habana-container-runtime
package is installed and set. For more details, refer to Installation Guide.
Installing MPI Operator¶
Follow MPI Operator documentation for instructions on setting up MPI Operator on your Kubernetes cluster.
Running Multi-Gaudi Workloads Example¶
Below is an example of a MPIJob on a MNIST model on 16 Gaudi devices.
Create
mpijob-mnist.yaml
file. Make sure to set the number of Gaudi nodes inWorker -> replicas
:apiVersion: kubeflow.org/v2beta1 kind: MPIJob metadata: name: mnist-run spec: slotsPerWorker: 8 runPolicy: cleanPodPolicy: Running mpiReplicaSpecs: Launcher: replicas: 1 template: spec: containers: - image: vault.habana.ai/gaudi-docker/1.16.2/ubuntu22.04/habanalabs/pytorch-installer-2.2.2:latest name: mnist-launcher command: ["/bin/bash", "-c"] args: - >- HOSTSFILE=$OMPI_MCA_orte_default_hostfile; MASTER_ADDR="$(head -n 1 $HOSTSFILE | sed -n s/[[:space:]]slots.*//p)"; NUM_NODES=$(wc -l < $HOSTSFILE); CARDS_PER_NODE=8; N_CARDS=$((NUM_NODES*CARDS_PER_NODE)); SETUP_CMD="git clone --branch 1.16.2 https://github.com/HabanaAI/Model-References.git /Model-References"; $SETUP_CMD; mpirun --npernode 1 \ --tag-output \ --allow-run-as-root \ --prefix $MPI_ROOT \ $SETUP_CMD; MODEL_PATH=/Model-References/PyTorch/examples/computer_vision/hello_world; MNIST_CMD="python $MODEL_PATH/mnist.py \ --batch-size=64 \ --epochs=1 \ --lr=1.0 \ --gamma=0.7 \ --hpu"; cd $MODEL_PATH; mpirun -np ${N_CARDS} \ --allow-run-as-root \ --bind-to core \ --map-by ppr:4:socket:PE=6 \ -rank-by core --report-bindings \ --tag-output \ --merge-stderr-to-stdout --prefix $MPI_ROOT \ -x MASTER_ADDR=$MASTER_ADDR \ $MNIST_CMD; Worker: replicas: 2 template: spec: hostIPC: true containers: - image: vault.habana.ai/gaudi-docker/1.16.2/ubuntu22.04/habanalabs/pytorch-installer-2.2.2:latest name: mnist-worker resources: limits: habana.ai/gaudi: 8 memory: 409Gi hugepages-2Mi: 95000Mi requests: habana.ai/gaudi: 8 memory: 409Gi hugepages-2Mi: 95000Mi
Note
PyTorch uses shared memory buffers to communicate between processes. By default, Docker containers are allocated 64MB of shared memory. When using more than one HPU, this allocation can be insufficient. Setting
hostIPC: true
allows re-using the host’s shared memory space inside the container.According to Kubernetes’ backoff policy, if a failure occurs, such as the worker pods are not running, the job is automatically restarted. This is useful for resuming long-running training from a checkpoint if an error causes the job to crash. For more information, refer to Kubernetes backoff failure policy.
Run the job:
kubectl apply -f mpijob-mnist.yaml
Check the pod status:
kubectl get pods -A
Retrieve the name of the pod and see the results:
kubectl logs <pod-name>