Skip to main content
EnterpriseAI Platform

Pre-Activation Checklist

Before you begin, make sure all of the following conditions are met. Every item is required:

RequirementDescription
HAMiThe cluster has been deployed and is running normally, and HAMi Enterprise / HAMi AI Platform has been installed
kubectl accessThe role associated with the current kubeconfig has permission to create and update secrets resources in the hami-system namespace

HAMi AI Platform Edition

Request a License​

Obtain the Certificate​

Complete the installation tasks and make sure all component Pods have started successfully before beginning the activation process.

  1. Log in to HAMi AI Platform with a Platform Administrator account.

  2. Go to the License and System Information page.

  3. Follow the on-screen instructions to obtain the license request information.

  4. Send the license request information to Dynamia AI sales or delivery personnel.

  5. Complete activation according to the guidance provided.

Receive the License​

After Dynamia AI issues the license, the sales team will deliver the License file by email.

Activate the License​

  1. Log in to HAMi AI Platform with a Platform Administrator account.

  2. Go to the License and System Information page.

  3. Click Activate License and select the License file received in the sales email.

  4. Click Confirm.

  5. Wait for the system to complete validation.

Check Scheduler Events​

View the Kubernetes events for secret/hami-license.

kubectl get events --field-selector involvedObject.name=hami-license -n hami-system

Expected output:

LAST SEEN   TYPE     REASON         OBJECT                MESSAGE
3m49s       Normal   LicenseValid   secret/hami-license   license c90a9410-5f94-44fd-95b8-XXXXXXXX validated successfully in secret hami-system/hami-license: expire_time=2027-06-18 08:01:51.85 +0000 UTC

Validation Results​

Please confirm the following:

  • The License status is displayed as "Normal".

  • The product and authorization scope match the purchased items.

  • The License expiration time is correct.

  • Platform features can be used normally.

HAMi Enterprise Edition

Receive the License​

After Dynamia AI issues the license, the sales team will deliver the License file by email.

Save the file to an operations machine that can access the target Kubernetes cluster through kubectl.

Create the License Secret​

Run the following command to create a Kubernetes Secret from the License file:

kubectl create secret generic hami-license \
  --from-file=license=/path/to/your License file \
  -n hami-system

Expected output:

secret/hami-license created

Add the License Label​

After the Secret is created successfully, run:

kubectl label secret hami-license \
  hami.io/license="true" \
  -n hami-system

Expected output:

secret/hami-license labeled

At this point, the HAMi Enterprise License has been deployed successfully.

Check the Secret​

Run:

kubectl get secret hami-license -n hami-system

If hami-license is returned, the Secret has been created successfully.

Check the License label:

kubectl get secret hami-license \
  -n hami-system \
  -o jsonpath='{.metadata.labels.hami\.io/license}'

Expected output:

true

Check Scheduler Events​

View the Kubernetes events for secret/hami-license.

kubectl get events --field-selector involvedObject.name=hami-license -n hami-system

Expected output:

LAST SEEN   TYPE     REASON         OBJECT                MESSAGE
3m49s       Normal   LicenseValid   secret/hami-license   license c90a9410-5f94-44fd-95b8-XXXXXXXX validated successfully in secret hami-system/hami-license: expire_time=2027-06-18 08:01:51.85 +0000 UTC

Verify Node License Registration​

Run the following command to check the License registration status of each GPU node:

$ kubectl get nodes -o custom-columns='NODE:.metadata.name,LICENSE:.metadata.annotations.hami\.io/nvidia-license'
NODE                                         LICENSE
gke-demo-gpu-pool-a843f02e-3fln   [{"uuid":"GPU-f9cd036c-382f-b13d-443f-8baac323f800","reminder":10,"expire":"2026-12-22T10:46:16.755Z"}]
gke-demo-gpu-pool-a843f02e-kbql   [{"uuid":"GPU-ba9a9a2d-0eda-cce5-0ac4-f961eb8020fd","reminder":10,"expire":"2026-12-22T10:46:16.755Z"}]
gke-demo-gpu-pool-a843f02e-n02f   [{"uuid":"GPU-daaff9d7-781c-db2b-f011-f6c28d99303d","reminder":10,"expire":"2026-12-22T10:46:16.755Z"}]

Field descriptions:

  • uuid: GPU device UUID. It should match the hardware information you provided to Dynamia AI.

  • expire: License expiration time (UTC)

Validation checklist (verify each item):

  • All GPU nodes appear in the list, and the LICENSE field is not empty.

  • The GPU UUID of each node matches the purchase list / environment information.

  • The License expiration time matches the agreed issuance period.

If any of the above does not match expectations, contact Dynamia AI after-sales / technical support immediately.

Example output (screenshot):

image

The License activation process is now complete.

Validation

Run a Sample Workload​

Deploy the following sample Pod to confirm that GPU resources can be scheduled and used normally by HAMi:

apiVersion: v1
kind: Pod
metadata:
  name: gpu-pod
spec:
  containers:
    - name: ubuntu-container
      image: ubuntu:22.04
      command: ["bash", "-c", "sleep 86400"]
      resources:
        limits:
          nvidia.com/gpu: 1 # declare how many physical GPUs the pod needs
          nvidia.com/gpumem: 3000 # identifies 3000M GPU memory each physical GPU allocates to the pod （Optional,Integer）
          nvidia.com/gpucores: 30 # identifies 30% GPU GPU core each physical GPU allocates to the pod （Optional,Integer)

Deploy and verify:

# Deploy the sample Pod
kubectl apply -f gpu-pod.yaml

# Wait until the Pod enters the Running state
kubectl get pod gpu-pod -w

# Enter the container and run nvidia-smi to confirm GPU resources are visible
kubectl exec -it gpu-pod -- nvidia-smi

Expected result: the Pod is in the Running state, and nvidia-smi can see the allocated GPU resources inside the container. The visible GPU memory and compute capacity are constrained by gpumem / gpucores.

License Update or Renewal

HAMi AI Platform Edition​

After receiving a new License file, return to the platform License page, upload it, and activate it again.

HAMi Enterprise Edition​

If the hami-license Secret already exists in the cluster, delete the existing Secret first:

kubectl delete secret hami-license -n hami-system

Then recreate it using the new License file:

kubectl create secret generic hami-license \
  --from-file=license=/actual/path/to/the/License/file \
  -n hami-system

Add the label again:

kubectl label secret hami-license \
  hami.io/license="true" \
  -n hami-system

Check the scheduler events:

View the Kubernetes events for secret/hami-license.

kubectl get events --field-selector involvedObject.name=hami-license -n hami-system

Expected output:

LAST SEEN   TYPE     REASON         OBJECT                MESSAGE
3m49s       Normal   LicenseValid   secret/hami-license   license c90a9410-5f94-44fd-95b8-XXXXXXXX validated successfully in secret hami-system/hami-license: expire_time=2027-06-18 08:01:51.85 +0000 UTC

Usually no restart is required after the update. Wait about 60 seconds and then verify node registration again.

FAQ

What happens if the License expires?​

  • Existing AI workloads that are already running are not affected to preserve business continuity

  • New scheduling requests will no longer be accepted, and existing services that restart will no longer be schedulable

  • Recovery : replace the Secret, and it will reactivate automatically within 60 seconds

Can one License be used for multiple clusters?​

  • No. Each License is bound to exactly one cluster.

How do I renew in an offline air-gapped environment?​

  1. The SE delivers the new license file through the customer channel (email / USB drive)

  2. The customer activates it again

  3. No restart is required. It takes effect within 60 seconds, with no Internet connectivity at all

Get Support

If you encounter any issues during activation or validation, contact Dynamia AI through the following channels:

If you encounter issues such as license exceptions, unregistered GPU nodes, or failed License updates, contact the Dynamia AI technical support team.