Which EMR Storage and Instance Choices Meet Cost and Reliability Needs?

DEA-C01 By islever Editorial Last updated

Automate data processing by using AWS services. Choose a data store.

Short answer

BD Use Amazon S3 as the persistent data store and Graviton instances for core and task nodes to lower EMR costs while maintaining Spark performance.

Question

A company is planning to use a provisioned Amazon EMR cluster that runs Apache Spark jobs to perform big data analysis. The company requires high reliability. A big data team must follow best practices for running cost-optimized and long-running workloads on Amazon EMR. The team must find a solution that will maintain the company's current level of performance. Which combination of resources will meet these requirements MOST cost-effectively? (Choose two.)
  1. Use Hadoop Distributed File System (HDFS) as a persistent data store.
  2. Use Amazon S3 as a persistent data store. Correct Answer
  3. Use x86-based instances for core nodes and task nodes.
  4. Use Graviton instances for core nodes and task nodes. Correct Answer
  5. Use Spot Instances for all primary nodes.

Key Takeaways

This question tests EMR storage durability and instance-family cost optimization; the trap is treating HDFS as persistent or putting Spot Instances on all primary nodes.

This DEA-C01 question tests cost-optimized, reliable Amazon EMR design for long-running Apache Spark workloads. The page establishes that Amazon S3 for persistent storage and Graviton instances for core and task nodes are the correct pair (B, D).

Common mistake: A common mistake is selecting HDFS (A) as the persistent store; EMR HDFS is tied to the cluster lifecycle and disappears on termination, whereas S3 provides durable storage decoupled from compute.

Expert Analysis

Why the Answer Is Correct

Amazon S3 is the preferred persistent data store for Amazon EMR because it decouples storage from the cluster lifecycle and provides durable, highly available storage for Spark input and output. Using S3 as the persistent store prevents the data loss that would occur if the cluster were terminated and leaves HDFS available only as scratch space. Graviton instances are ARM-based instances that AWS positions as lower-cost alternatives to comparable x86 instances while maintaining performance for many Spark workloads, so core and task nodes can be run more cost-effectively. Together, S3 and Graviton satisfy the reliability requirement without sacrificing the current performance level.

Why the Other Options Are Wrong

HDFS on EMR is ephemeral and is not suitable as a persistent store because the data is lost when the cluster terminates; long-running jobs can also exhaust local disk. x86-based core and task instances can work, but they are generally less cost-effective than Graviton when performance is comparable. Using Spot Instances for all primary nodes undermines the high-reliability requirement because Spot capacity can be reclaimed and interrupt core or master workloads. These alternatives either add cost, reduce durability, or introduce availability risk.

Community Comment Notes

Several learners converge on BD, and one explains that "HDFS is not recommended for persistent storage" because cluster termination destroys HDFS data. Another commenter captures the AWS-native rule of thumb with "Graviton is aws processor, and also EMRFS on S3." A dissenting answer of BE appears, but it conflicts with the high-reliability requirement because Spot for all primary nodes risks interruption. The consensus supports S3 for durable storage and Graviton for lower-cost compute.

Official Reference

These links open the official vendor documentation in a new tab.

Exam Strategy

For EMR cost questions, separate storage durability from compute purchasing: persistent data should live in S3, and Graviton is the default lower-cost instance family when performance must be maintained. Eliminate options that put Spot on primary nodes or use HDFS as durable storage before comparing the remaining pair.

Frequently Asked Questions

Why is HDFS (A) not suitable as a persistent store for long-running EMR Spark jobs?

HDFS on EMR is ephemeral: data is lost when the cluster terminates, and long-running jobs can fill local disk. Amazon S3 decouples storage from compute and provides durable, scalable persistence.

Why choose Graviton instances (D) instead of x86 instances for EMR core and task nodes?

Graviton instances are ARM-based and AWS positions them as lower-cost alternatives to comparable x86 instances while maintaining performance for many Spark workloads.

More DEA-C01 FAQ →

Community Votes

BD
100%

100% of anonymous learners picked answer BD. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 7 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 7

  1. [Removed] 👍 9
    HDFS is not recommended for persistent storage because once a cluster is terminated, all HDFS data is lost. Also, long-running workloads can fill the disk space quickly. Thus, S3 is the best option since it's highly available, durable, and scalable. AWS Graviton-based instances cost up to 20% less than comparable x86-based Amazon EC2 instances: https://aws.amazon.com/ec2/graviton/
  2. pypelyncar 👍 3
    s3 no question. Graviton=> Cost-Effectiveness: Graviton instances are ARM-based instances specifically designed for cloud workloads. They offer significant cost savings compared to x86-based instances while delivering comparable or better performance for many Apache Spark workloads. Performance: Graviton instances are optimized for Spark workloads and can deliver the same level of performance as x86-based instances in many cases. Additionally, EMR offers performance-optimized versions of Spark built for Graviton instances.
  3. certplan 👍 2
    A. - AWS recommends using Amazon S3 as a persistent data store for Amazon EMR due to its scalability, durability, and cost-effectiveness. Storing data in HDFS would require managing and maintaining additional infrastructure, which may incur higher costs in terms of storage, management, and scalability compared to using Amazon S3. AWS documentation emphasizes the benefits of integrating Amazon EMR with Amazon S3 for cost optimization and efficiency. D. - While Graviton instances may offer cost savings in certain scenarios, they might not always be the most cost-effective option depending on the specific workload requirements and availability of compatible software. x86-based instances are more commonly supported by a broader range of software and frameworks, which could result in better performance and compatibility in some cases. Additionally, AWS documentation on instance types and pricing can provide insights into the cost-effectiveness of Graviton instances compared to x86-based instances.
  4. ttpro1995 👍 1
    Rule of thumb: pick the AWS in-house solution provided for that service. Graviton is aws processor, and also EMRFS on S3.
  5. sam_pre 👍 1
    Cost effective + high reliability > S3 Gravitation > Low cost
Show 2 more comments
  1. GiorgioGss 👍 3
    B and D.
  2. okechi 👍 1
    My answer is BE

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

Amazon Redshift Compound Sort Key Query Optimization

S-grade

Secure Redshift Spectrum Access with VPC Endpoints

S-grade

Athena Federated Query for Multi-Source Iceberg Pipelines

S-grade

Reclaiming Storage in Amazon Redshift Materialized Views

S-grade

Practice All DEA-C01 Questions

Access 100 questions with complete answers and detailed explanations.

View Full DEA-C01 Practice Test →

← Back to DEA-C01 Study Guide