Skip to content

deploying nim on k8s. With this custom-value.yaml, 3.1 8b model can be deployed but 70b failed. #92

Description

@SarielMa

we follow the steps here: https://docs.nvidia.com/nim/large-language-models/latest/deploy-helm.html

after helm install ....

kubectl logs my-nim-01 --previous

...
{"level": "None", "time": "None", "file_name": "None", "file_path": "None", "line_number": "-1", "message": "[09-25 19:03:45.989 ERROR nim_sdk::hub::repo rust/nim-sdk/src/hub/repo.rs:119] error sending request for url (https://api.ngc.nvidia.com/v2/org/nim/team/meta/models/llama-3_1-70b-instruct/hf-1d54af3-nim1.2/files)", "exc_info": "None", "stack_info": "None"}
{"level": "ERROR", "time": "None", "file_name": "None", "file_path": "None", "line_number": "-1", "message": "", "exc_info": "Traceback (most recent call last):\n File "/usr/lib/python3.10/runpy.py", line 196, in _run_module_as_main\n return _run_code(code, main_globals, None,\n File "/usr/lib/python3.10/runpy.py", line 86, in _run_code\n exec(code, run_globals)\n File "/opt/nim/llm/vllm_nvext/entrypoints/launch.py", line 99, in \n main()\n File "/opt/nim/llm/vllm_nvext/entrypoints/launch.py", line 42, in main\n inference_env = prepare_environment()\n File "/opt/nim/llm/vllm_nvext/entrypoints/args.py", line 155, in prepare_environment\n engine_args, extracted_name = inject_ngc_hub(engine_args)\n File "/opt/nim/llm/vllm_nvext/hub/ngc_injector.py", line 247, in inject_ngc_hub\n cached = repo.get_all()\nException: error sending request for url (https://api.ngc.nvidia.com/v2/org/nim/team/meta/models/llama-3_1-70b-instruct/hf-1d54af3-nim1.2/files)", "stack_info": "None"}

kubectl describe pod my-nim-01
...
Events:
Type Reason Age From Message


Warning BackOff 5m3s (x90 over 102m) kubelet Back-off restarting failed container nim-llm in pod my-nim-0_default(ce8f1e3a-f0e6-4a95-9086-2901091b7a57)
Normal Pulled 4m52s (x15 over 116m) kubelet Container image "nvcr.io/nim/meta/llama-3.1-70b-instruct:latest" already present on machine

kubectl get pods -A
NAMESPACE NAME READY STATUS RESTARTS AGE
default my-nim-0 0/1 Running 14 (6m46s ago) 117m

vim custom-value.yaml
image:
repository: "nvcr.io/nim/meta/llama-3.1-70b-instruct" # container location
tag: latest # NIM version you want to deploy
model:
ngcAPISecret: ngc-api # name of a secret in the cluster that includes a key named NGC_API_KEY and is an NGC API key
imagePullSecrets:

  • name: ngc-secret # name of a secret used to pull nvcr.io images, see https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
    persistence:
    enabled: true
    size: 800Gi
    accessMode: ReadWriteMany
    storageClass: ""
    annotations:
    helm.sh/resource-policy: "keep"
    livenessProbe:
    initialDelaySeconds: 600
    periodSeconds: 60
    timeoutSeconds: 10
    startupProbe:
    initialDelaySeconds: 600
    periodSeconds: 60
    timeoutSeconds: 10
    failureThreshold: 1500
    resources:
    limits:
    nvidia.com/gpu: 4 # much more GPU memory is required

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions