What are Amazon EC2 Spot Instances?
Amazon EC2 Spot Instances let you use spare EC2 capacity at steeply reduced prices. AWS states savings of up to 90 percent compared with On-Demand prices. In exchange, Amazon EC2 can reclaim the instance as soon as the capacity is needed elsewhere. AWS calls this event a Spot Instance interruption.
AWS describes Spot Instances as designed for stateless, fault-tolerant or flexible applications. There is no bidding: the Spot price is derived dynamically from available capacity.
Core capabilities
- Significant price advantage: Up to 90 percent off On-Demand prices
- Interruption notice: Warning two minutes before stop or termination, delivered through EventBridge and instance metadata
- Selectable behavior: On interruption the instance is terminated, stopped or hibernated
- No bidding: Dynamic pricing based on available capacity
- Broad integration: Usable through EC2 Auto Scaling, Amazon ECS, Amazon EKS, Amazon EMR and Amazon SageMaker
- Optional maximum price: AWS notes that setting a maximum price leads to more frequent interruptions
Typical use cases
Container workloads: Amazon ECS, Amazon EKS and self-managed Kubernetes clusters use Spot capacity for node groups that tolerate interruptions.
Big data and machine learning: Amazon EMR, Hadoop, Spark and Amazon SageMaker process batch and training jobs that can resume or repeat after an interruption.
HPC and simulation: Compute-intensive tasks such as genomic sequencing or algorithmic trading are spread across many parallel, interruptible instances.
CI/CD pipelines: Build and test agents run on Spot capacity, and interrupted runs are restarted.
Benefits
- Substantial cost reduction for interruptible workloads
- Access to spare EC2 capacity without a bidding process
- Two minutes of warning for a controlled shutdown
- Integration with existing AWS capacity management services
Integration with innFactory
As an AWS Reseller, innFactory supports you with Amazon EC2 Spot Instances: assessing which workloads are suitable, designing interruption-tolerant architectures, configuring Auto Scaling groups and EKS node groups, and evaluating the savings achieved.
Typical Use Cases
Technical Specifications
Frequently Asked Questions
What are Amazon EC2 Spot Instances?
Amazon EC2 Spot Instances use spare EC2 capacity at significantly reduced rates. AWS states savings of up to 90 percent compared with On-Demand prices. In exchange, AWS can reclaim the instances when the capacity is needed again.
How much warning is there before an interruption?
AWS describes the Spot Instance interruption notice as a warning issued two minutes before the instance is stopped or terminated. If hibernation is set as the interruption behavior, you receive a notice but no two-minute warning, because the hibernation process begins immediately.
Which workloads suit Spot Instances?
AWS describes Spot Instances as designed for stateless, fault-tolerant or flexible applications. Typical examples are container workloads, big data and machine learning jobs, HPC applications and CI/CD pipelines.
How do I learn about an upcoming interruption?
Interruption notices are made available as an Amazon EventBridge event and as an item in the instance metadata. AWS recommends checking the metadata every five seconds and notes that notices are emitted on a best-effort basis.
Which AWS services work with Spot Instances?
AWS names EC2 Auto Scaling, Amazon ECS, Amazon EKS, Amazon EMR and Amazon SageMaker, as well as tools such as Jenkins and Spinnaker. Spot capacity is launched, scaled and managed through them.
Note: All product information on this page has been compiled with care, but is provided without guarantee and may be outdated or incomplete. Cloud services evolve rapidly — features, pricing, SLAs, and availability change frequently. Authoritative and up-to-date information can only be found on the official product page of AWS (official documentation). This page does not represent an offer by AWS.