Cost reduction with Amazon EMR on EKS
January 06, 2025
Cost reduction with Amazon EMR on EKS

Score 8 out of 10
Vetted Review
Verified User
Overall Satisfaction with Amazon EMR (Elastic MapReduce)
Amazon EMR (Elastic MapReduce) is heavily used at my organization for most if not all data pipeline computations: we started by using EC2 instances, we then moved to EMR Serverless and we are actually completing the transition to EMR on EKS. In general we use it for long-running analysis (SQLs with a lot of JOINs) and overall for batch processing. From what I've seen, we use it with Spark under the hood.
Pros
- EMR on EKS is really flexible and cost-saving
- Flexibility on how to run the jobs (and different implementations to choose from)
- Support online and it's a regularly updated product
Cons
- EMR on EKS could be better documented, especially since for the "magic" it does under the hood when using Spark
- UI can be improved (especially for EMR on EKS)
- Switching to EMR on EKS most of our EMR on EC2 jobs has produced a reduction of 4% in the overall costs (while maintaining the same level of data freshness)
Do you think Amazon EMR (Elastic MapReduce) delivers good value for the price?
Yes
Are you happy with Amazon EMR (Elastic MapReduce)'s feature set?
Yes
Did Amazon EMR (Elastic MapReduce) live up to sales and marketing promises?
I wasn't involved with the selection/purchase process
Did implementation of Amazon EMR (Elastic MapReduce) go as expected?
Yes
Would you buy Amazon EMR (Elastic MapReduce) again?
Yes
Comments
Please log in to join the conversation