Skip to content

Pre-requisites for Deploying NAI

In this part of the lab we will prepare pre-requisites for LLM application on GPU nodes.

The following is the flow of the applications lab:

stateDiagram-v2
    direction LR

    state PreRequisites {
        [*] --> CreateFilesShare  
        CreateFilesShare --> PrepareHuggingFace
        PrepareHuggingFace --> [*]
    }
    state CreateOfflineHelmContainers {
        [*] --> PrepareNAIHelmCharts
        PrepareNAIHelmCharts --> PrepareNAIContainerImages
        PrepareNAIContainerImages --> [*]
    }

    [*] --> PreRequisites
    PreRequisites --> CreateOfflineHelmContainers
    CreateOfflineHelmContainers --> DeployNAI : next section
    DeployNAI --> TestNAI
    TestNAI --> [*]

Prepare the following pre-requisites needed to deploy NAI on target kubernetes cluster.

Create Nutanix Files Storage Class

We will create Nutanix Files storage class which will be used to create a pvc that will store the LLama-3-8B model files.

  1. In Prism Central, choose Files from the menu
  2. Choose the file server (e.g. labFS)
  3. Click on Shares & Exports
  4. Click on +New Share or Export

  5. Fill the details of the Share

    • Name - model_share
    • Description - for NAI model store
    • Share path - leave blank
    • Max Size - 10 GiB (adjust to the model file size)
    • Primary Protocol Access - NFS

  6. Click Next and make sure Enable compression in checked

  7. Click Next

  8. In NFS Protocol Access, choose the following:

    • Authentication - System
    • Default Access (for all clients) - Read-Write
    • Squash - Root Squash

    Note

    Consider changing access options for Production environment

  9. Click Next

  10. Confirm the share details and click on Create

Create the Files Storage Class

  1. Run the following command to check K8S status of the nkpdev cluster

    kubectl get nodes

  2. In VSC Explorer, click on New File and create a config file with the following name:

    nai-nfs-storage.yaml

    Add the following content and replace the nfsServerName with the name of the Nutanix Files server name .

    Finding nfsServerName and nfsServer fqdn

    apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: nai-nfs-storage
parameters:
  nfsPath: <nfs-path>
  nfsServer: <nfs-server>
  storageType: NutanixFiles
provisioner: csi.nutanix.com
reclaimPolicy: Delete
volumeBindingMode: Immediate

    apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  name: nai-nfs-storage
parameters:
  nfsPath: /model_share
  nfsServer: labFS.ntnxlab.local
  storageType: NutanixFiles
provisioner: csi.nutanix.com
reclaimPolicy: Delete
volumeBindingMode: Immediate

  3. Create the storage class

    kubectl apply -f nai-nfs-storage.yaml

  4. Check storage classes in the cluster for the Nutanix Files storage class

    kubectl get storageclass

    kubectl get storageclass

NAME                       PROVISIONER                     RECLAIMPOLICY   VOLUMEBINDINGMODE      ALLOWVOLUMEEXPANSION   AGE
dkp-object-store           kommander.ceph.rook.io/bucket   Delete          Immediate              false                  28h
nai-nfs-storage            csi.nutanix.com                 Delete          Immediate              true                   24h
nutanix-volume (default)   csi.nutanix.com                 Delete          WaitForFirstConsumer   false                  28h

Request Access to Model on Hugging Face

Follow these steps to request access to the meta-llama/Meta-Llama-3.1-8B-Instruct model:

LLM Recommendation

From testing google/gemma-2-2b-it model is quicker to download and obtain download rights, than meta-llama/Meta-Llama-3.1-8B-Instruct model.

Feel free to use the google/gemma-2-2b-it model if necessary. The procedure to request access to the model is the same.

  1. Sign in to your Hugging Face account:

  2. Navigate to the model page:

  3. Request access:

    • On the model page, you will see a section or button labeled Request Access (this is usually near the top of the page or near the "Files and versions" section).
    • Click Request Access.

  4. Complete the form:

    • You may be prompted to fill out a form or provide additional details about your intended use of the model.
    • Complete the required fields and submit the request.

  5. Wait for approval:

    • After submitting your request, you will receive a notification or email once your access is granted.
    • This process can take some time depending on the approval workflow.

Once access is granted, there will be an email notification.

Note

Email from Hugging Face can take a few minutes or hours before it arrives.

Create a Hugging Face Token with Read Permissions

Follow these steps to create a Hugging Face token with read permissions:

  1. Sign in to your Hugging Face account:

  2. Access your account settings:

    • Click on your profile picture in the top-right corner.
    • From the dropdown, select Settings.

  3. Navigate to the "Access Tokens" section:

    • In the sidebar, click on Access Tokens.
    • You will see a page where you can create and manage tokens.

  4. Create a new token:

    • Click the New token button.
    • Enter a name for your token (i.e., read-only-token).

  5. Set token permissions:

    • Under the permissions dropdown, select Read. For Example: hf-token

  6. Create and copy the token:

    • After selecting the permissions, click Create.
    • Your token will be generated and displayed only once, so make sure to copy it and store it securely.

Use this token for accessing Hugging Face resources with read-only permissions.

Prepare Helm Charts

In this section we will prepare the helm charts necessary for NAI and pre-requisite applications install

  • NAI
  • Envoy Gateway
  • Kserve
  • OpenTelemetry Operator

The procedure will be done on the jumphost VM.

  1. Login to Nutanix Portal using your credentials

  2. Go to Downloads > NAI Airgapped Bundle

  3. Download and extract the NAI air-gap helm bundle from the link you copied earlier

    Paste the download URL within double quotes
    curl -o nai-helm-charts-2.5.0.tar "_paste_download_URL_here"

    curl -o nai-helm-charts-2.5.0.tar "https://download.nutanix.com/downloads/nai/2.5.0/nai-core-2.5.0.tar?........"
    1. Open new VSCode window on your jumphost VM

  4. In VSCode Explorer pane, click on existing $HOME folder

  5. Click on New Folder name it: airgap-nai

  6. On VSCode Explorer plane, click the $HOME/airgap-nai folder

  7. On VSCode menu, select Terminal > New Terminal

  8. Browse to airgap-nai directory

    cd $HOME/airgap-nai

  9. In VSC, under the newly created airgap-nai folder, click on New File and create file with the following name:

    .env

  10. Add (append) the following environment variables and save it

    export REGISTRY_HOST=harbor.10.x.x.111.nip.io/nkp
export REGISTRY_USERNAME=admin
export REGISTRY_CACERT=_path_to_ca_cert_of_registry  # (1)!
    1. File must contain CA server and Harbor server's public certificate in one file
    export REGISTRY_USERNAME=admin
export REGISTRY_PASSWORD=xxxxxxx
export REGISTRY_CACERT=$HOME/harbor/certs/full_chain.pem  # (1)!
    1. File must contain CA server and Harbor server's public certificate in one file

  11. Source the .env file to import environment variables

    source $HOME/airgap-nai/.env

  12. Download the NAI 2.5.0 helm chart bundle from Nutanix Portal

    curl -o "nai-helm-charts-2.5.0.tar" "_paste_download_URL_here"

    curl -o "nai-helm-charts-2.5.0.tar" "https://download.nutanix.com/downloads/nai/2.5.0/nai-helm-charts-2.5.0.tar?Expires=xxxxx"

  13. Login to Harbor registry on the command line (if not done so)

    helm registry login harbor.10.x.x.111.nip.io --username ${REGISTRY_USERNAME} --password ${REGISTRY_PASSWORD}

  14. Upload the downloaded and prepared helm charts to Harbor

    # Push charts
for chart in $(ls *.tgz); do echo $chart;helm push $chart oci://$REGISTRY_HOST;done

    for chart in $(ls *.tgz); do echo $chart;helm push $chart oci://$REGISTRY_HOST;done
gateway-crds-helm-v1.5.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/gateway-crds-helm:v1.5.0
Digest: sha256:d8f06d0d28bbcb40188f211be0fc3f0dee20376c5ad59935bf68bc5cc67149fc
gateway-helm-v1.5.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/gateway-helm:v1.5.0
Digest: sha256:4e49511296e23e3d1400c92cfb38a5c26030501ec7353883e4ccad9fd7cc4c2c
kserve-crd-v0.15.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/kserve-crd:v0.15.0
Digest: sha256:d540337971ae91fbd7593764eb90d8689012bc835f5d2f8ee42af47543a18403
kserve-v0.15.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/kserve:v0.15.0
Digest: sha256:cafd90ab1d91a54a28c1ff2761d976bdda0bb173675ef392a16ac250b044d15f
nai-core-2.5.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/nai-core:2.5.0
Digest: sha256:6326c6348f31c2ee0de93d03b164863addb82f8054283a4d418c8d2910b7935d
nai-operators-2.5.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/nai-operators:2.5.0
Digest: sha256:09775842c1bae10a094f54a54bc3f82177f0a9b92f0e52a20219d1cc7bf41d96
opentelemetry-operator-0.93.0.tgz
Pushed: harbor.10.x.x.134.nip.io/nkp/opentelemetry-operator:0.93.0
Digest: sha256:01e050311f4f335fb2b4ae9c3dabe6efc7017cfba37cc3dc8c9acce756e52732
Now the charts are available in the OCI compatible container/chart registry.

Prepare Container Images

The Jumphost VM will be used as a medium to download the NAI container images and upload them to the internal Harbor container registry.

stateDiagram-v2
    direction LR

    state LoginToNutanixPortal {
        [*] --> CreateDockerIDandAccessToken
        CreateDockerIDandAccessToken --> LoginToDockerCLI
        LoginToDockerCLI --> [*]
    }

    state PrepareNAIDockerImages {
        [*] --> DownloadUploadImagesToHarbor
        DownloadUploadImagesToHarbor --> [*]
    }

    [*] --> LoginToNutanixPortal
    LoginToNutanixPortal --> PrepareNAIDockerImages
    PrepareNAIDockerImages --> [*]

Upload NAI Docker Images to Harbor

Info

The download and upload of the container images will be done in one docker push command which will use the internal Harbor container registry details.

This will be a two-step process.

  1. Upload the container images from the downloaded nai-2.x.x.tar to the jumphost VM local docker images store
  2. Upload it to the internal Harbor container registry

  1. Download the NAI air-gap bundles (NAI container images) from the Nutanix Portal > Downloads > Nutanix Enterprise AI

    Paste the download URL within double quotes
    curl -o nai-2.5.0-1.tar "_paste_download_URL_here"
curl -o nai-2.5.0-1.tar "_paste_download_URL_here"
    This download is about 44 GBs
    curl -o nai-2.5.0-1.tar "https://download.nutanix.com/downloads/nai/2.5.0/nai-2.5.0-1.tar?..."
curl -o nai-2.5.0-2.tar "https://download.nutanix.com/downloads/nai/2.5.0/nai-2.5.0-2.tar?..."

  2. Since we will be using the same internal Harbor container registry to upload container images, make sure the following environment variables are set (these were already set during air-gap NKP preparation)

    export REGISTRY_HOST=harbor.10.x.x.111.nip.io/nkp
export REGISTRY_USERNAME=admin
export REGISTRY_PASSWORD=xxxxxxx
export REGISTRY_CACERT=$HOME/harbor/certs/ca.crt

  3. (Optional) - To view the container images loaded in your local docker container registry, run the following command:

    docker images --format '{{.Repository}}:{{.Tag}}' | grep nai

    nutanix/nai-api:v2.5.0
nutanix/nai-inference-ui:v2.5.0
nutanix/nai-model-processor:v2.5.0
nutanix/nai-iep-operator:v2.5.0
nutanix/nai-tgi:3.3.4-b2485c9
nutanix/nai-kserve-huggingfaceserver:v0.15.2
nutanix/nai-kserve-huggingfaceserver:v0.15.2
nutanix/nai-kserve-huggingfaceserver:v0.15.2-gpu
nutanix/nai-kserve-controller:v0.15.0
nutanix/nai-postgres:16.1-alpine

  4. Push the images to the jumphost VM local docker images store

    docker image load -i nai-2.5.0-1.tar
docker image load -i nai-2.5.0-2.tar

  5. Tag and push all the NAI images to refer to the internal harbor registry

    for image in $(docker images --format '{{.Repository}}:{{.Tag}}' | grep 'nai' | grep -v ${REGISTRY_HOST}); do
  docker tag $image ${REGISTRY_HOST}/$(echo $image);
  docker push ${REGISTRY_HOST}/$(echo $image);
done

Now we are ready to deploy our NAI workloads.