Closed didumin closed 3 months ago
Hello, be sure that the model was actually deployed and the container is healthy.
Hello, @bsekachev, model status in nuclio is running
container details also shows "Status: Running"
% kubectl describe pods -n cvat nuclio-pod-name-7f46bdfbfb-4pqml Name: nuclio-pod-name-7f46bdfbfb-4pqml Namespace: cvat Priority: 0 Service Account: default Node: cl1751vimeosmurmvbkr-eqid/10.2.0.11 Start Time: Fri, 05 Apr 2024 15:22:07 +0300 Labels: nuclio.io/app=functionres nuclio.io/class=function nuclio.io/function-name=pod-name nuclio.io/function-version=latest nuclio.io/project-name=cvat pod-template-hash=7f46bdfbfb Annotations: framework: onnx name: cvat-ui-model-name nuclio.io/image-hash: 1712319722245667772 spec: [ { "id": 0, "name": "pricetag" }, { "id": 1, "name": "price" } ] Status: Running IP: 10.112.133.151 IPs: IP: 10.112.133.151 Controlled By: ReplicaSet/nuclio-pod-name-7f46bdfbfb Containers: nuclio: Container ID: containerd://b2d440f3b5281b942f959bfc379ef517e7d50bf1c86e65825252f75e31b94ab5 Image: reg.gitlab.itnap.ru/machine-learning/infrastructure_and_tools/cvat_autolabeling/cvat.onnx:latest Image ID: reg.gitlab.itnap.ru/machine-learning/infrastructure_and_tools/cvat_autolabeling/cvat.onnx@sha256:046d44fdf48f22b0acf1256bcfe1af6a4e3b6ad7b0aa584dc0f09030c81edafe Port: 8080/TCP Host Port: 0/TCP State: Running Started: Fri, 05 Apr 2024 15:22:08 +0300 Ready: True Restart Count: 0 Requests: cpu: 25m memory: 1Mi Liveness: http-get http://:8082/live delay=10s timeout=3s period=5s #success=1 #failure=3 Readiness: http-get http://:8080/__internal/health delay=5s timeout=1s period=1s #success=1 #failure=10 Environment: FUNCTION_TASK_ID: b90731540cb94742af4fe1f85fae4f2b FUNCTION_LABELS: [ { "id": 0, "name": "pricetag" }, { "id": 1, "name": "price" } ]
NUCLIO_FUNCTION_NAME: pod-name
NUCLIO_FUNCTION_VERSION: latest
NUCLIO_FUNCTION_INSTANCE: nuclio-pod-name-7f46bdfbfb-4pqml (v1:metadata.name)
Mounts:
/etc/nuclio/config/platform from platform-config-volume (rw)
/etc/nuclio/config/processor from processor-config-volume (rw)
/var/run/secrets/kubernetes.io/serviceaccount from kube-api-access-wjkxq (ro)
Conditions:
Type Status
Initialized True
Ready True
ContainersReady True
PodScheduled True
Volumes:
platform-config-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: nuclio-platform-config
Optional: true
processor-config-volume:
Type: ConfigMap (a volume populated by a ConfigMap)
Name: nuclio-pod-name
Optional: false
kube-api-access-wjkxq:
Type: Projected (a volume that contains injected data from multiple sources)
TokenExpirationSeconds: 3607
ConfigMapName: kube-root-ca.crt
ConfigMapOptional:
Maybe there are any additional troubleshooting steps or some kind of info how cvat do discover nuclio models?
@bsekachev may you suggest any further investigation steps or head me to related docs?
Actions before raising this issue
Steps to Reproduce
helm chart cvat deployed according to https://docs.cvat.ai/docs/administration/advanced/k8s_deployment_with_helm/ nuclio enabled and deployed as well after deploying model via nuclio dashboard UI a pod with model exist
Expected Behavior
model appears in cvat
Possible Solution
No response
Context
no models in cvat models tab
Environment