Request a Quota Increase
Your users might use Amazon resources in amounts that exceed those specified by their quotas. If your users are resource constrained and encounter errors in SageMaker Canvas, you can request a quota increase for them.
For more details about SageMaker quotas and how to request a quota increase, see Quotas.
Amazon SageMaker Canvas uses the following services to process the requests of your users:
-
Amazon SageMaker Autopilot
-
Amazon SageMaker Studio Classic domain
-
Amazon Forecast
For a list of the available quotas for SageMaker Canvas operations that aren't used to forecast time series data, see Amazon SageMaker endpoints and quotas.
For a list of the available quotas for SageMaker Canvas operations that are used to forecast time series data, see Amazon Forecast endpoints and quotas.
Request an increase for instances to build custom models
When building a custom model, if you encounter an error during post-building analysis that
tells you to increase your quota for ml.m5.2xlarge
instances, use the following
information to resolve the issue.
You must increase the SageMaker
Hosting endpoint quota for the ml.m5.2xlarge
instance type to a non-zero value in
your Amazon account. After building a model, SageMaker Canvas hosts the model on a SageMaker Hosting endpoint and
uses the endpoint to generate the post-building analysis. If you don't increase the default
account quota of 0 for ml.m5.2xlarge
instances, SageMaker Canvas cannot complete this step and
generates an error during post-building analysis.
For the procedure to increase the quota, see Requesting a quota increase in the Service Quotas User Guide.