AWS Sagemaker request quota for training job usage

When requesting quota increases for SageMaker training job usage, the approval process isn't always instant and can vary based on the specific instance type and region.

While some quota increase requests might be approved automatically, others require manual review by AWS Support, which appears to be happening in your case. This is normal, especially for specialized instance types like ml.f5.xlarge.

The fact that a support ticket was created indicates your request is being reviewed. The approval timeline can vary from a few hours to a few days, depending on the request complexity and current AWS capacity in the eu-central-1 region.

If your request is urgent for your upcoming conference, I recommend:

1. Adding a comment to your existing support ticket explaining the urgency and your timeline
2. Checking the status of your request in the Service Quotas console under the "Request history" tab
3. If you don't receive a response within 24 hours, consider contacting AWS Support directly through the support center to escalate your request

For future reference, when you encounter a training job limit error, it typically appears as a "ResourceLimitExceeded" message indicating you've exceeded the number of training jobs you can run with that instance type. The quota increase process is the correct solution for this issue.

It's worth noting that deployment endpoints and training jobs have different quota systems, which explains why your deployment quota request might have been approved instantly while the training job request requires review.

Sources
View SageMaker training plans quotas using the AWS management console - Amazon SageMaker AI
Troubleshooting guide - Amazon SageMaker AI