Home icon

Introducing Amazon EMR Serverless Flex: Run non-urgent Spark jobs at a discount

Big Data Blog



This article introduces Amazon EMR Serverless Flex, a new execution option for running non-urgent Spark batch jobs on spare compute capacity at a 60% discount.

  • Flex handles spare capacity best practices automatically: AZ selection, instance diversification, driver protection, and decommission-aware scheduling
  • Spark driver runs on dedicated Standard worker; executors run on discounted Flex workers with automatic task retry on reclamation
  • Benchmark shows 63% cost savings with acceptable runtime tradeoff (7-30% slower depending on scale)
  • Flex and Standard jobs run on same application; enable with single parameter in job submission
  • Recommended practices: avoid top-of-hour peaks, run outside business hours, use Serverless Storage for shuffle-heavy jobs

EMR Serverless Flex enables customers to consolidate entire Spark platforms onto serverless infrastructure by supporting cost-optimized non-urgent workloads previously requiring EC2 Spot Instances.



Go to article

The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.

Related articles

Nov 17
2026
Introducing the Apache Spark Optimization Agent for Amazon EMR Serverless
Nov 18
2026
Introducing the Apache Spark optimization agent for Amazon EMR Serverless and AWS Glue ETL
Jun 9
2026
Run Interactive Workloads on Amazon EMR Serverless with Spark Connect
Nov 21
2025
Amazon EMR Serverless now supports Apache Spark 4.0.1 (preview)

The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.