System Requirements

The Run:ai Cluster is a Kubernetes application.

This article explains the required hardware and software system requirements for the Run:ai cluster.

Set out below are the system requirements for the Run:ai cluster.

Hardware Requirements¶

The following hardware requirements are for the Kubernetes Cluster nodes’. By default, all Run:ai cluster services run on all available nodes. For production deployments, you may want to Set Node Roles, to separate between system and worker nodes, reduce downtime and save CPU cycles on expensive GPU Machines.

Run:ai Cluster - system nodes¶

This configuration is the minimum requirement you need to install and use Run:ai Cluster.

Run:ai Cluster - Worker nodes¶

Run:ai Workloads require compute resources, Run:ai Cluster supports both x86 CPU and NVIDIA x86 GPUs.

The following GPU models are supported:

• NVIDIA H100
• NVIDIA A100-80
• NVIDIA A100-40
• NVIDIA A30
• NVIDIA Tesla V100
• NVIDIA Tesla T4

Shared storage¶

Run:ai workloads must be able to access data from any worker node in a uniform way, to access training data and code as well as save checkpoints, weights, and other machine-learning-related artifacts.

Typical protocols are Network File Storage (NFS) or Network-attached storage (NAS). Run:ai Cluster supports both, for more information see Shared storage.

Software requirements¶

The following software requirements must be fulfilled on the Kubernetes cluster.

Operating system¶

• Any Linux operating system supported by both Kubernetes and NVIDIA GPU Operator
• Run:ai cluster on Google Kubernetes Engine (GKE) only supports Ubuntu, as Container-Optimized OS (COS) is not supported by NVIDIA GPU Operator
• Internal tests are being performed on Ubuntu 20.04 and CoreOS for OpenShift.

Kubernetes distribution¶

Run:ai Cluster requires Kubernetes. The following Kubernetes distributions are supported:

• Vanilla Kubernetes
• OpenShift Container Platform (OCP)
• NVIDIA Base Command Manager (BCM)
• Elastic Kubernetes Engine (EKS)
• Google Kubernetes Engine (GKE)
• Azure Kubernetes Service (AKS)
• Rancher Kubernetes Engine (RKE1)
• Rancher Kubernetes Engine 2 (RKE2)

Contact Run:ai customer support for up-to-date support details for:

• Ezmeral Runtime Enterprise
• Tanzu Platform

For existing Kubernetes clusters, see the following Kubernetes version support matrix for the latest Run:ai cluster releases:

For information on supported versions of managed Kubernetes, it's important to consult the release notes provided by your Kubernetes service provider. There, you can confirm the specific version of the underlying Kubernetes platform supported by the provider, ensuring compatibility with Run:ai. For an up-to-date end-of-life statement see Kubernetes Release History or OpenShift Container Platform Life Cycle Policy

Kubernetes Pod Security Admission¶

Run:ai v2.15 and above supports restricted policy for Pod Security Admission (PSA) on OpenShift only. Other Kubernetes distributions are only supported with privileged policy.

For Run:ai on OpenShift to run with PSA restricted policy:

• Label the runai namespace as described in Pod Security Admission with the following labels:

pod-security.kubernetes.io/audit=privileged
pod-security.kubernetes.io/enforce=privileged
pod-security.kubernetes.io/warn=privileged

• The workloads submitted through Run:ai should comply with the restrictions of PSA restricted policy, This can be enforced using Policies.

Kubernetes Ingress Controller¶

Run:ai cluster requires Kubernetes Ingress Controller to be installed on the Kubernetes cluster.

• OpenShift, RKE and RKE2 come pre-installed ingress controller.
• Internal tests are being performed on NGINX, Rancher NGINX, OpenShift Router, and Istio.
• Make sure that a default ingress controller is set.

There are many ways to install and configure different ingress controllers. A simple example to install and configure NGINX ingress controller using helm:

helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo update
helm upgrade -i nginx-ingress ingress-nginx/ingress-nginx \
    --namespace nginx-ingress --create-namespace \
    --set controller.kind=DaemonSet \
    --set controller.service.externalIPs="{<INTERNAL-IP>,<EXTERNAL-IP>}" # Replace <INTERNAL-IP> and <EXTERNAL-IP> with the internal and external IP addresses of one of the nodes

helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo update
helm install nginx-ingress ingress-nginx/ingress-nginx \
    --namespace nginx-ingress --create-namespace

NVIDIA GPU Operator¶

Run:ai Cluster requires NVIDIA GPU Operator to be installed on the Kubernetes Cluster, supports version 22.9 to 23.9

See the Installing the NVIDIA GPU Operator, followed by notes below:

• Use the default gpu-operator namespace . Otherwise, you must specify the target namespace using the flag runai-operator.config.nvidiaDcgmExporter.namespace as described in customized cluster installation.
• NVIDIA drivers may already be installed on the nodes. In such cases, use the NVIDIA GPU Operator flags --set driver.enabled=false. DGX OS is one such example as it comes bundled with NVIDIA Drivers.
• To use Dynamic MIG, the GPU Operator must be installed with the flag mig.strategy=mixed as described in customized cluster installation. If the GPU Operator is already installed, edit the clusterPolicy by running

kubectl patch clusterPolicy cluster-policy -n gpu-operator --type=merge -p '{"spec":{"mig":{"strategy": "mixed"}}}

• For distribution-specific additional instructions see below:

kubectl create ns gpu-operator

apiVersion: v1
kind: ResourceQuota
metadata:
name: gcp-critical-pods
namespace: gpu-operator
spec:
scopeSelector:
    matchExpressions:
    - operator: In
    scopeName: PriorityClass
    values:
    - system-node-critical
    - system-cluster-critical

kubectl apply -f resourcequota.yaml

For troubleshooting information, see the NVIDIA GPU Operator Troubleshooting Guide.

Prometheus¶

Run:ai Cluster requires Prometheus to be installed on the Kubernetes cluster.

• OpenShift comes pre-installed with prometheus
• For RKE2 see Enable Monitoring instructions to install Prometheus

There are many ways to install Prometheus. A simple example to install the community Kube-Prometheus Stack using helm, run the following commands:

helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo update
helm install prometheus prometheus-community/kube-prometheus-stack \
    -n monitoring --create-namespace --set grafana.enabled=false

Optional software requirements¶

Optional Run:ai capabilities, Distributed Training and Inference require additional Kubernetes applications (frameworks) to be installed on the cluster.

Distributed training¶

Distributed training enables training of AI models over multiple nodes. This requires distributed training framework to be installed on the cluster. The following frameworks are supported:

• TensorFlow
• PyTorch
• XGBoost
• MPI v2

There are many ways to install each framework. A simple example of installation is the Kubeflow Training Operator. Run:ai supports Kubeflow version 1.7 only - which includes TensorFlow, Pytorch, and XGBoost.

To install run the following command:

kubectl apply -k "github.com/kubeflow/training-operator.git/manifests/overlays/standalone?ref=v1.7.0"

To install MPI v2, which is not included in the Kubeflow Training Operator, run the following command:

kubectl apply --server-side -f https://raw.githubusercontent.com/kubeflow/mpi-operator/master/deploy/v2beta1/mpi-operator.yaml

kubectl patch deployment training-operator -n kubeflow --type='json' -p='[{"op": "add", "path": "/spec/template/spec/containers/0/args", "value": ["--enable-scheme=tfjob", "--enable-scheme=pytorchjob", "--enable-scheme=xgboostjob"]}]'
kubectl delete crd mpijobs.kubeflow.org

Inference¶

Inference enables serving of AI models. This requires the Knative Serving framework to be installed on the cluster and supports Knative versions 1.4 to 1.12

Follow the Installing Knative instructions. After installation, configure Knative to use the Run:ai scheduler and features, by running:

kubectl patch configmap/config-autoscaler \
  --namespace knative-serving \
  --type merge \
  --patch '{"data":{"enable-scale-to-zero":"true"}}' && \
kubectl patch configmap/config-features \
  --namespace knative-serving \
  --type merge \
  --patch '{"data":{"kubernetes.podspec-schedulername":"enabled","kubernetes.podspec-affinity":"enabled","kubernetes.podspec-tolerations":"enabled","kubernetes.podspec-volumes-emptydir":"enabled","kubernetes.podspec-securitycontext":"enabled","kubernetes.podspec-persistent-volume-claim":"enabled","kubernetes.podspec-persistent-volume-write":"enabled","multi-container":"enabled","kubernetes.podspec-init-containers":"enabled"}}'

Domain Name Requirement¶

The following requirement must be followed for naming the domain.

Fully Qualified Domain Name (FQDN)¶

You must have a Fully Qualified Domain Name (FQDN) to install Run:ai Cluster (ex: runai.mycorp.local). This cannot be an IP. The domain name must be accessible inside the organization only. You also need a TLS certificate (private and public) for HTTPS access.