ML Delete Idle Sagemaker Models

Opportunity Name:

ML Delete Idle SageMaker Models

 

AWS Resource Type:

SageMaker Model

 

Opportunity Description:

SageMaker models that have been idle for an extended period can incur unnecessary costs. CloudFix identifies these idle models, which have not been associated with any endpoint or transform jobs in the last 30 days, and recommends their deletion to save costs.

 

Criteria for identifying the opportunity:

• Model must be a SageMaker Model.
• Annual cost, extrapolated from the last 31 days of usage, exceeds the annual public cost threshold (default $100).
• Model has not been associated with any endpoint configurations.
• Model has not been used in any transform jobs in the last 30 days.

 

Potential savings (range in % on annual basis):

• Savings are calculated based on the cost of running the idle SageMaker model that has not been used for the last 30 days. The exact savings will depend on the specific costs associated with the identified idle model.

 

Can CloudFix implement the fix automatically once I accept the recommendation?

No

 

Other considerations:

• Performance impact: Deleting idle SageMaker models has no impact on active models or workloads.
• Data loss considerations: Ensure that the model is truly idle and not expected to be used in the near future before deletion to avoid any unintended loss of model data.
• Security concerns: Review the security policies and permissions related to the SageMaker models to ensure that deleting an idle model does not affect other dependent resources or workflows.

 

Additional Resources:

• CloudFix Knowledge Base
• AWS SageMaker DeleteModel API