question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Unable to serve on AWS S3

See original GitHub issue

/kind bug

What steps did you take and what happened:

I want to serve a pytorch model using kf serving on AWS (i am following this tutorial: https://github.com/kubeflow/kfserving/tree/master/docs/samples/v1alpha2/pytorch). Here are my steps:

s3_secret.yaml

apiVersion: v1
kind: Secret
metadata:
  name: mysecret
  annotations:
     serving.kubeflow.org/s3-endpoint: s3://kfserving-samples/models/pytorch/cifar10 # replace with your s3 endpoint
     serving.kubeflow.org/s3-usehttps: "1" # by default 1, for testing with minio you need to set to 0
type: Opaque
data:
  AWS_ACCESS_KEY_ID: bWluaW8=
  AWS_SECRET_ACCESS_KEY: bWluaW8xMjM=
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: sa
secrets:
  - name: mysecret
$ kubectl apply -f s3_secret.yaml

secret/mysecret configured
serviceaccount/sa configured

You then set the serviceAccountName on the KFService

pytorch_serve.yaml

apiVersion: "serving.kubeflow.org/v1alpha2"
kind: "InferenceService"
metadata:
  name: "pytorch-s3-cifar10-005"
spec:
  default:
    predictor:
      serviceAccountName: sa
      pytorch:
        storageUri: "s3://kfserving-samples/models/pytorch/cifar10"


$ kubectl apply -f pytorch_serve.yaml
inferenceservice.serving.kubeflow.org/pytorch-s3-cifar10-005 created

I dont get any URL generated in the output

$ kubectl get inferenceservice
NAME                     URL                                          READY   DEFAULT TRAFFIC   CANARY TRAFFIC   AGE

pytorch-s3-cifar10-005                                                False                                      6m15s

What did you expect to happen: $ kubectl get inferenceservice should have generated URL like: http://pytorch-cifar10-005.default.example.com

Anything else you would like to add: Am I missing any steps ? What should I do to make this work ?

Environment:

  • Kubeflow version: 1.2
  • Kubernetes version: (use kubectl version): 1.18.9
  • OS (e.g. from /etc/os-release): ubuntu 18.04

Issue Analytics

  • State:closed
  • Created 2 years ago
  • Comments:15 (6 by maintainers)

github_iconTop GitHub Comments

1reaction
ajinkya933commented, Apr 5, 2021

I fixed it corect SERVICE_HOSTNAME=$(kubectl get inferenceservice s3-test-004 -o jsonpath=‘{.status.url}’ | cut -d “/” -f 3)

Thanks for all your patient efforts and apt followups 💯

0reactions
juanvasquezreyescommented, Jun 3, 2021

I just figured it out, I was using the wrong region, but I’m still getting having issues when using tls #1630

Read more comments on GitHub >

github_iconTop Results From Across the Web

Troubleshooting - Amazon Simple Storage Service
If an invalid ACL is specified or bucket ACL permissions grant access outside of your AWS account, you might see the following error...
Read more >
Troubleshoot endpoint URL connection error in Amazon S3
I'm trying to run the cp or sync command on my Amazon Simple Storage Service (Amazon S3) bucket. However, I'm getting the "Could...
Read more >
I can't route traffic to an Amazon S3 bucket that's configured for ...
When you configure an Amazon S3 bucket for website hosting, you must give the bucket the same name as the record that you...
Read more >
Resolve "Unable to locate credentials" error in Amazon S3
An "Unable to locate credentials" error indicates that Amazon S3 can't find the credentials to authenticate AWS API calls. To resolve this issue ......
Read more >
I can't connect to Amazon S3 - Amazon QuickSight
In the https://console.aws.amazon.com/s3/ , navigate to your Amazon S3 bucket, choose the Permissions tab, and add the appropriate permissions. Make sure that ...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found