This document is relevant for: Inf1, Inf2, Trn1, Trn1n

Deploy Containers with Neuron#

In this section you will find resources to help you use containers for your accelerated deep learning model acceleration on top of Inferentia and Trainium enabled instances.

The section is organized based on the target deployment environment and use case. In most cases, it is recommended to use a preconfigured Deep Learning Container (DLC) from AWS. Each DLC is pre-configured to have all of the Neuron components installed and is specific to the chosen ML Framework.

Locate Neuron DLC image

Find latest Neuron DLC images.

In the DLC release page do a search for Neuron to get the ECR repo location of specific Neuron DLC release.

Getting Started
Launch Trn1 Instance
  • Please follow the instructions at launch an Amazon EC2 Instance to Launch an instance, when choosing the instance type at the EC2 console. Please make sure to select the correct instance type.

  • To get more information about instances sizes and pricing see: Trn1 web page, Inf2 web page, Inf1 web page

  • Select your Amazon Machine Image (AMI) of choice, please note that Neuron supports Amazon Linux 2 AMI(HVM) - Kernel 5.10.

  • When launching a Trn1, please adjust your primary EBS volume size to a minimum of 512GB.

  • After launching the instance, follow the instructions in Connect to your instance to connect to the instance

Note

If you are facing a connectivity issue during the model loading process on a Trn1 instance with Ubuntu, that could probably be because of Ubuntu limitations with multiple interfaces. To solve this problem, please follow the steps mentioned here.

Users are highly encouraged to use DLAMI to launch the instances, since DLAMIs come with the required fix.

Install Drivers
# Configure Linux for Neuron repository updates

sudo tee /etc/yum.repos.d/neuron.repo > /dev/null <<EOF
[neuron]
name=Neuron YUM Repository
baseurl=https://yum.repos.neuron.amazonaws.com
enabled=1
metadata_expire=0
EOF
sudo rpm --import https://yum.repos.neuron.amazonaws.com/GPG-PUB-KEY-AMAZON-AWS-NEURON.PUB

# Update OS packages
sudo yum update -y

# Install OS headers
sudo yum install kernel-devel-$(uname -r) kernel-headers-$(uname -r) -y

# Remove preinstalled packages and Install Neuron Driver and Runtime
sudo yum remove aws-neuron-dkms -y
sudo yum remove aws-neuronx-dkms -y
sudo yum install aws-neuronx-dkms-2.*  -y

# Install EFA Driver(only required for multi-instance training)
curl -O https://efa-installer.amazonaws.com/aws-efa-installer-latest.tar.gz
wget https://efa-installer.amazonaws.com/aws-efa-installer.key && gpg --import aws-efa-installer.key
cat aws-efa-installer.key | gpg --fingerprint
wget https://efa-installer.amazonaws.com/aws-efa-installer-latest.tar.gz.sig && gpg --verify ./aws-efa-installer-latest.tar.gz.sig
tar -xvf aws-efa-installer-latest.tar.gz
cd aws-efa-installer && sudo bash efa_installer.sh --yes
cd
sudo rm -rf aws-efa-installer-latest.tar.gz aws-efa-installer
Install Docker
sudo yum install -y docker.io
sudo usermod -aG docker $USER

Logout and log back in to refresh membership.
Verify Docker
docker run hello-world

Expected result:

Hello from Docker!
This message shows that your installation appears to be working correctly.

To generate this message, Docker took the following steps:
1. The Docker client contacted the Docker daemon.
2. The Docker daemon pulled the "hello-world" image from the Docker Hub.
(amd64)
3. The Docker daemon created a new container from that image which runs the
executable that produces the output you are currently reading.
4. The Docker daemon streamed that output to the Docker client, which sent it
to your terminal.

To try something more ambitious, you can run an Ubuntu container with:
$ docker run -it ubuntu bash

Share images, automate workflows, and more with a free Docker ID:
https://hub.docker.com/

For more examples and ideas, visit:
https://docs.docker.com/get-started/
Verify Neuron Component

Once the environment is setup, a container can be started with –device=/dev/neuron# to specify desired set of Inferentia/Trainium devices to be exposed to the container. To find out the available neuron devices on your instance, use the command ls /dev/neuron*.

When running neuron-ls inside a container, you will only see the set of exposed Trainiums. For example:

docker run --device=/dev/neuron0 neuron-test neuron-ls

Would produce the following output in trn1.32xlarge:

::

   +--------+--------+--------+---------+
   | NEURON | NEURON | NEURON |   PCI   |
   | DEVICE | CORES  | MEMORY |   BDF   |
   +--------+--------+--------+---------+
   | 0      | 2      | 32 GB  | 10:1c.0 |
   +--------+--------+--------+---------+
Build and Run Docker Image
Run Tutorial
Launch Inf1 Instance
Install Drivers
# Configure Linux for Neuron repository updates
sudo tee /etc/yum.repos.d/neuron.repo > /dev/null <<EOF
[neuron]
name=Neuron YUM Repository
baseurl=https://yum.repos.neuron.amazonaws.com
enabled=1
metadata_expire=0
EOF
sudo rpm --import https://yum.repos.neuron.amazonaws.com/GPG-PUB-KEY-AMAZON-AWS-NEURON.PUB

# Update OS packages
sudo yum update -y

################################################################################################################
# To install or update to Neuron versions 1.19.1 and newer from previous releases:
# - DO NOT skip 'aws-neuron-dkms' install or upgrade step, you MUST install or upgrade to latest Neuron driver
################################################################################################################

# Install OS headers
sudo yum install kernel-devel-$(uname -r) kernel-headers-$(uname -r) -y

# Install Neuron Driver
sudo yum install aws-neuron-dkms -y

####################################################################################
# Warning: If Linux kernel is updated as a result of OS package update
#          Neuron driver (aws-neuron-dkms) should be re-installed after reboot
####################################################################################
Install Docker
sudo yum install -y docker.io
sudo usermod -aG docker $USER

Logout and log back in to refresh membership.
Verify Docker
docker run hello-world

Expected result:

Hello from Docker!
This message shows that your installation appears to be working correctly.

To generate this message, Docker took the following steps:
1. The Docker client contacted the Docker daemon.
2. The Docker daemon pulled the "hello-world" image from the Docker Hub.
(amd64)
3. The Docker daemon created a new container from that image which runs the
executable that produces the output you are currently reading.
4. The Docker daemon streamed that output to the Docker client, which sent it
to your terminal.

To try something more ambitious, you can run an Ubuntu container with:
$ docker run -it ubuntu bash

Share images, automate workflows, and more with a free Docker ID:
https://hub.docker.com/

For more examples and ideas, visit:
https://docs.docker.com/get-started/
Verify Neuron Component

Once the environment is setup, a container can be started with –device=/dev/neuron# to specify desired set of Inferentia/Trainium devices to be exposed to the container. To find out the available neuron devices on your instance, use the command ls /dev/neuron*.

When running neuron-ls inside a container, you will only see the set of exposed Inferentias. For example:

docker run --device=/dev/neuron0 neuron-test neuron-ls

Would produce the following output in inf1.xlarge:

::

   +--------------+---------+--------+-----------+-----------+------+------+
   |   PCI BDF    | LOGICAL | NEURON |  MEMORY   |  MEMORY   | EAST | WEST |
   |              |   ID    | CORES  | CHANNEL 0 | CHANNEL 1 |      |      |
   +--------------+---------+--------+-----------+-----------+------+------+
   | 0000:00:1f.0 |       0 |      4 | 4096 MB   | 4096 MB   |    0 |    0 |
   +--------------+---------+--------+-----------+-----------+------+------+
Run Tutorial
Kubernetes Getting Started
Prerequisite
  • Working kubernetes cluster

  • Inf1/Trn1 instances as worker nodes with attached roles allowing: * ECR read access policy to retrieve container images from ECR: arn:aws:iam::aws:policy/AmazonEC2ContainerRegistryReadOnly

  • Tutorial Docker environment setup: to install required packages in the worker nodes. With EKS, the EKS optimized accelarated AMI has the necessary neuron components installed

  • Kubernetes node object has instance-type set to inf1/trn1 types. For ex, "node.kubernetes.io/instance-type": "inf1.2xlarge" or "node.kubernetes.io/instance-type": "trn1.2xlarge"

Deploy Neuron Device Plugin

Neuron device plugin exposes Neuron cores & devices to kubernetes as a resource. aws.amazon.com/neuroncore, aws.amazon.com/neurondevice, aws.amazon.com/neuron are the resources that the neuron device plugin registers with the kubernetes. aws.amazon.com/neuroncore is used for allocating neuron cores to the container. aws.amazon.com/neurondevice is used for allocating neuron devices to the container. When neurondevice is used all the cores belonging to the device will be allocated to container. aws.amazon.com/neuron also allocates neurondevices and this exists just to be backward compatible with already existing installations. aws.amazon.com/neurondevice is the recommended resource for allocating devices to the container.

  • Make sure prequisite are satisified

  • Download the neuron device plugin yaml file. k8s-neuron-device-plugin.yml

  • Download the neuron device plugin rbac yaml file. This enables permissions for device plugin to update the node and Pod annotations. k8s-neuron-device-plugin-rbac.yml

  • Apply the Neuron device plugin as a daemonset on the cluster with the following command

    kubectl apply -f k8s-neuron-device-plugin-rbac.yml
    kubectl apply -f k8s-neuron-device-plugin.yml
    
  • Verify that neuron device plugin is running

    kubectl get ds neuron-device-plugin-daemonset --namespace kube-system
    

    Expected result (with 2 nodes in cluster):

    NAME                             DESIRED   CURRENT   READY   UP-TO-DATE   AVAILABLE   NODE SELECTOR   AGE
    neuron-device-plugin-daemonset   2         2         2       2            2           <none>          27h
    
  • Verify that the node has allocatable neuron cores and devices with the following command

    kubectl get nodes "-o=custom-columns=NAME:.metadata.name,NeuronCore:.status.allocatable.aws\.amazon\.com/neuroncore"
    

    Expected result:

    NAME                                          NeuronCore
    ip-192-168-65-41.us-west-2.compute.internal   32
    ip-192-168-87-81.us-west-2.compute.internal   32
    
    kubectl get nodes "-o=custom-columns=NAME:.metadata.name,NeuronDevice:.status.allocatable.aws\.amazon\.com/neurondevice"
    

    Expected result:

    NAME                                          NeuronDevice
    ip-192-168-65-41.us-west-2.compute.internal   16
    ip-192-168-87-81.us-west-2.compute.internal   16
    
Deploy Neuron Scheduler Extension

Neuron scheduler extension is required for scheduling pods that require more than one Neuron core or device resource. Refer Neuron Scheduler Extension Flow Diagram for details on how the neuron scheduler extension works. Neuron scheduler extension filter out nodes with non-contiguous core/device ids and enforces allocation of contiguous core/device ids for the PODs requiring it.

In cluster environments where there is no access to default scheduler, the neuron scheduler extension can be used with another scheduler. A new scheduler is added (along with the default scheduler) and then the pod’s that needs to run the neuron workload use this new scheduler. Neuron scheduler extension is added to this new scheduler. EKS natively does not yet support the neuron scheduler extension and so in the EKS environment this is the only way to add the neuron scheduler extension.

  • Make sure Neuron device plugin is running

  • Download the my scheduler my-scheduler.yml

  • Download the scheduler extension k8s-neuron-scheduler-eks.yml

  • Apply the neuron-scheduler-extension

    kubectl apply -f k8s-neuron-scheduler-eks.yml
    
  • Apply the my-scheduler.yml

    kubectl apply -f my-scheduler.yml
    
  • Check there are no errors in the my-scheduler pod logs and the k8s-neuron-scheduler pod is bound to a node

    kubectl logs -n kube-system my-scheduler-79bd4cb788-hq2sq
    
    I1012 15:30:21.629611       1 scheduler.go:604] "Successfully bound pod to node" pod="kube-system/k8s-neuron-scheduler-5d9d9d7988-xcpqm" node="ip-192-168-2-25.ec2.internal" evaluatedNodes=1 feasibleNodes=1
    
  • When running new pod’s that need to use the neuron scheduler extension, make sure it uses the my-scheduler as the scheduler. Sample pod spec is below

    apiVersion: v1
    kind: Pod
    metadata:
    name: <POD_NAME>
    spec:
    restartPolicy: Never
    schedulerName: my-scheduler
    containers:
        - name: <POD_NAME>
        command: ["<COMMAND>"]
        image: <IMAGE_NAME>
        resources:
            limits:
            cpu: "4"
            memory: 4Gi
            aws.amazon.com/neuroncore: 9
            requests:
            cpu: "1"
            memory: 1Gi
    
  • Once the neuron workload pod is run, make sure logs in the k8s neuron scheduler has successfull filter/bind request

    kubectl logs -n kube-system k8s-neuron-scheduler-5d9d9d7988-xcpqm
    
    2022/10/12 15:41:16 POD nrt-test-5038 fits in Node:ip-192-168-2-25.ec2.internal
    2022/10/12 15:41:16 Filtered nodes: [ip-192-168-2-25.ec2.internal]
    2022/10/12 15:41:16 Failed nodes: map[]
    2022/10/12 15:41:16 Finished Processing Filter Request...
    
    2022/10/12 15:41:16 Executing Bind Request!
    2022/10/12 15:41:16 Determine if the pod %v is NeuronDevice podnrt-test-5038
    2022/10/12 15:41:16 Updating POD Annotation with alloc devices!
    2022/10/12 15:41:16 Return aws.amazon.com/neuroncore
    2022/10/12 15:41:16 neuronDevUsageMap for resource:aws.amazon.com/neuroncore in node: ip-192-168-2-25.ec2.internal is [false false false false false false false false false false false false false false false false]
    2022/10/12 15:41:16 Allocated ids for POD nrt-test-5038 are: 0,1,2,3,4,5,6,7,8
    2022/10/12 15:41:16 Try to bind pod nrt-test-5038 in default namespace to node ip-192-168-2-25.ec2.internal with &Binding{ObjectMeta:{nrt-test-5038    8da590b1-30bc-4335-b7e7-fe574f4f5538  0 0001-01-01 00:00:00 +0000 UTC <nil> <nil> map[] map[] [] []  []},Target:ObjectReference{Kind:Node,Namespace:,Name:ip-192-168-2-25.ec2.internal,UID:,APIVersion:,ResourceVersion:,FieldPath:,},}
    2022/10/12 15:41:16 Updating the DevUsageMap since the bind is successful!
    2022/10/12 15:41:16 Return aws.amazon.com/neuroncore
    2022/10/12 15:41:16 neuronDevUsageMap for resource:aws.amazon.com/neuroncore in node: ip-192-168-2-25.ec2.internal is [false false false false false false false false false false false false false false false false]
    2022/10/12 15:41:16 neuronDevUsageMap for resource:aws.amazon.com/neurondevice in node: ip-192-168-2-25.ec2.internal is [false false false false]
    2022/10/12 15:41:16 Allocated devices list 0,1,2,3,4,5,6,7,8 for resource aws.amazon.com/neuroncore
    2022/10/12 15:41:16 Allocated devices list [0] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [0] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [0] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [0] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [1] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [1] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [1] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [1] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Allocated devices list [2] for other resource aws.amazon.com/neurondevice
    2022/10/12 15:41:16 Return aws.amazon.com/neuroncore
    2022/10/12 15:41:16 Succesfully updated the DevUsageMap [true true true true true true true true true false false false false false false false]  and otherDevUsageMap [true true true false] after alloc for node ip-192-168-2-25.ec2.internal
    2022/10/12 15:41:16 Finished executing Bind Request...
    
  • Make sure Neuron device plugin is running

  • Download the scheduler config map k8s-neuron-scheduler-configmap.yml

  • Download the scheduler extension k8s-neuron-scheduler.yml

  • Enable the kube-scheduler with option to use configMap for scheduler policy. In your cluster.yml Please update the spec section with the following

    spec:
      kubeScheduler:
      usePolicyConfigMap: true
    
  • Launch the cluster

    kops create -f cluster.yml
    kops create secret --name neuron-test-1.k8s.local sshpublickey admin -i ~/.ssh/id_rsa.pub
    kops update cluster --name neuron-test-1.k8s.local --yes
    
  • Apply the k8s-neuron-scheduler-configmap.yml [Registers neuron-scheduler-extension with kube-scheduler]

    kubectl apply -f k8s-neuron-scheduler-configmap.yml
    
  • Launch the neuron-scheduler-extension

    kubectl apply -f k8s-neuron-scheduler.yml
    
Tutorials
Developer Flows
FAQ, Troubleshooting and Release Note

This document is relevant for: Inf1, Inf2, Trn1, Trn1n