MLS-C01 Question 39
Select 3You are a machine learning engineer tasked with processing a large dataset to extract features for a recommendation system. The dataset is stored in Amazon S3, and the processing involves multiple steps such as data cleaning, transformation, and aggregation. You decide to use Amazon EMR to manage the process. Which of the following steps should you take to ensure efficient processing and cost optimization when using Amazon EMR?
- A
Choose a Spot Instance fleet for the core and task nodes to reduce costs.
- B
Enable EMRFS consistent view to ensure strong consistency when reading and writing data to Amazon S3.
- C
Use the latest version of Apache Spark available in Amazon EMR to benefit from performance enhancements.
- D
Configure data compression for intermediate and final output files to optimize storage and network usage.
- E
Provision a single master node with on-demand pricing to ensure cluster stability.
Show answer and explanation
Correct answers: A, C, D
Explanation
To ensure efficient processing and cost optimization on Amazon EMR, it is important to leverage Spot Instances for cost savings, use the latest software versions for performance enhancements, and enable data compression to minimize storage and network costs. While consistent view and a stable master node are important considerations, they are not directly relevant to cost reduction and processing efficiency in this scenario, making the other options more appropriate.
- A. Correct.
Choosing a Spot Instance fleet for core and task nodes can significantly reduce costs, as Spot Instances are typically cheaper than On-Demand Instances. However, you need to ensure your workload can handle potential interruptions.
- B. Incorrect.
While enabling EMRFS consistent view provides strong consistency, it is not necessary for most use cases unless you are dealing with scenarios where Amazon S3's eventual consistency model could cause issues. This might add unnecessary latency and cost.
- C. Correct.
Using the latest version of Apache Spark in Amazon EMR ensures that you take advantage of the latest performance optimizations and features, improving processing efficiency.
- D. Correct.
Configuring data compression reduces the size of intermediate and final output files, optimizing storage in S3 and reducing data transfer costs and processing latency.
- E. Incorrect.
While using an On-Demand Instance for the master node ensures stability, this option alone does not contribute to cost optimization and efficiency, making it less relevant to the scenario.