Sagemaker canvas: Not able to run predict either batch prediction / single prediction on my model

0

Hi, I see sagemaker canvas consistently return below error when I try to execution a batch prediction or single prediction.

Request Failed:
Canvas can't process your request right now. Try again in a few minutes, or contact your administrator and share the details below to resolve the issue.
If you're an administrator or an individual user, contact AWS support and provide the following code: <client_request_id> to resolve the issue.

There are no errors in cloud watch logs as well that direct to the issue.

nm
asked 2 months ago123 views
2 Answers
0

Hello there @nm, have you made sure to add the right permissions to your SageMaker Execution Role? You need the following AWS managed policy, AmazonSageMakerCanvasFullAccess, or make sure that your execution role has at least permissions to execute the actions CreateTransformJob and DescribeTransformJob.

AWS
answered 2 months ago
  • Hi @dgallitelli, thank you so much for your response. I rechecked the access and my role does have AmazonSageMakerCanvasFullAccess policy.

0

This might be due to permission or quota issues from experience. You can follow these steps below.

1.** Check AWS Service Health Dashboard** First, check the AWS Service Health Dashboard to ensure there are no ongoing issues with Amazon SageMaker Canvas or related AWS services in your region. If there are any service disruptions or maintenance events, they could be the cause of your issue.

  1. Model State: Verify that your model is in a usable state. If the model was recently trained or updated, make sure the process completed successfully.

  2. Review SageMaker Canvas Quotas Amazon SageMaker Canvas has quotas for the number of models you can train, the size of the data you can use, and other resources. Check if you've reached any of these quotas, as hitting a limit could prevent further predictions. https://docs.aws.amazon.com/sagemaker/latest/dg/canvas-requesting-quota-increases.html

  3. If problem still persist, Check for Permissions Issues If you're using SageMaker Canvas within an organization, ensure you have the necessary permissions to run predictions. Sometimes, restrictions on resources or actions can lead to errors. Attach AmazonSageMakerCanvasFullAccess, to the SageMaker Execution Role. However, if you are using a custom IAM configuration, this will override SageMaker Domain Execution Policy, then you must explicitly add permissions to your user's IAM execution role. Otherwise make sure your IAM user has these permissions "sagemaker:CreateAutoMLJobV2", "sagemaker:DescribeAutoMLJobV2". Follow this instruction to do this, https://docs.aws.amazon.com/sagemaker/latest/dg/canvas-set-up-cv-nlp.html

  4. Contact AWS Support As the error message suggests, if the above steps don’t resolve your issue, contact AWS support. When contacting them, provide the <client_request_id> mentioned in the error message. This ID helps AWS support to trace the request in their systems and diagnose the problem more efficiently.

AWS
answered 18 days ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions