kleveross / ormb

Docker for Your ML/DL Models Based on OCI Artifacts
Apache License 2.0
461 stars 61 forks source link

SeldonDeployment 的pod没有ready #189

Open gongyu143 opened 3 years ago

gongyu143 commented 3 years ago

[root@xxxx ormb]# kubectl get all NAME READY STATUS RESTARTS AGE pod/tfserving-default-0-mnist-model-6cdfd5d-6vncx 0/2 Running 0 108m

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE service/kubernetes ClusterIP 10.96.0.1 443/TCP 3d21h service/tfserving-default ClusterIP 10.102.59.194 8000/TCP,5001/TCP 109m service/tfserving-default-mnist-model ClusterIP 10.109.93.101 9000/TCP 109m

NAME READY UP-TO-DATE AVAILABLE AGE deployment.apps/tfserving-default-0-mnist-model 0/1 1 0 108m

NAME DESIRED CURRENT READY AGE replicaset.apps/tfserving-default-0-mnist-model-6cdfd5d 1 1 0 108m

kubectl describe pod/tfserving-default-0-mnist-model-6cdfd5d-6vncx

Events: Type Reason Age From Message


Warning Unhealthy 40s (x1699 over 4h18m) kubelet Liveness probe errored: strconv.Atoi: parsing "http": invalid syntax

kubectl logs -f tfserving-default-0-mnist-model-6cdfd5d-6vncx mnist-model

2021-09-30 06:10:13.660649: W tensorflow_serving/sources/storage_path/file_system_storage_path_source.cc:268] No versions of servable mnist-model found under base path /mnt/models. Did you forget to name your leaf directory as a number (eg. '/1/')? 2021-09-30 06:10:14.660861: W tensorflow_serving/sources/storage_path/file_system_storage_path_source.cc:268] No versions of servable mnist-model found under base path /mnt/models. Did you forget to name your leaf directory as a number (eg. '/1/')? 2021-09-30 06:10:15.662074: W tensorflow_serving/sources/storage_path/file_system_storage_path_source.cc:268] No versions of servable mnist-model found under base path /mnt/models. Did you forget to name your leaf directory as a number (eg. '/1/')? 2021-09-30 06:10:16.662489: W tensorflow_serving/sources/storage_path/file_system_storage_path_source.cc:268] No versions of servable mnist-model found under base path /mnt/models. Did you forget to name your leaf directory as a number (eg. '/1/')?

kubectl logs -f tfserving-default-0-mnist-model-6cdfd5d-6vncx mnist-model-model-initializer

Using / as the root path Logging to the remote registry demo.goharbor.io Username: ormbtest Login insecurely time="2021-09-30T01:55:43Z" level=debug msg="attempting v2 login to registry endpoint https://demo.goharbor.io/v2/" Login succeeded Using / as the root path v1: Pulling from demo.goharbor.io/tensorflow/fashion_model ref: demo.goharbor.io/tensorflow/fashion_model:v1 digest: 62c7d273602a9dc6833a1401e0186f50708e1e03b717edc4b0b39781b53c367a size: 162.1 KiB format: SavedModel Status: Downloaded newer model for demo.goharbor.io/tensorflow/fashion_model:v1 ref: demo.goharbor.io/tensorflow/fashion_model:v1 digest: 62c7d273602a9dc6833a1401e0186f50708e1e03b717edc4b0b39781b53c367a size: 162.1 KiB format: SavedModel Moving /mnt/models/model/saved_model.pb to /mnt/models/saved_model.pb Moving /mnt/models/model/variables to /mnt/models/variables

有没有大神知道怎么解决

gaocegege commented 3 years ago

Maybe you can attach to the container and find if there is a model downloaded.