DEA-C01 Question 328
Select 4You are designing a data processing pipeline for a financial analytics application. The pipeline ingests streaming data from stock market feeds, processes it in near real-time, and stores the results in a data warehouse for further analysis. You must choose between using AWS Lambda with Amazon Kinesis Data Streams (serverless) or Amazon EC2 instances with Apache Kafka (provisioned) for data ingestion and processing. Which factors should you consider when deciding between these options?
- A
The expected volume and rate of incoming data streams
- B
The need for fine-grained control over the underlying infrastructure
- C
The ability to scale automatically based on demand
- D
The upfront cost and time required to set up the solution
- E
The ability to integrate with other AWS managed services like Amazon S3 and DynamoDB
Show answer and explanation
Correct answers: A, B, C, D
Explanation
Choosing between serverless and provisioned services depends on factors like expected data volume, control over infrastructure, scalability needs, and setup complexity. Serverless options like AWS Lambda excel in ease of use and automatic scaling but have limitations in throughput and control. Provisioned services like EC2 with Kafka offer higher control and can handle larger workloads but require more effort for setup and scaling. Integration with AWS managed services is typically supported by both approaches, so it is less critical in the decision-making process.
- A. Correct.
The volume and rate of incoming data streams are critical factors. Serverless services like AWS Lambda have limits on invocation rates, while provisioned services like EC2 with Kafka can handle higher throughputs but require manual scaling.
- B. Correct.
Provisioned services like EC2 with Kafka offer fine-grained control over the underlying infrastructure, which is useful for complex configurations, whereas serverless services abstract this control.
- C. Correct.
Serverless services like AWS Lambda automatically scale based on demand, whereas provisioned services like EC2 require manual or auto-scaling configurations.
- D. Correct.
Provisioned services like EC2 with Kafka require significant upfront setup, whereas serverless services like AWS Lambda can be deployed more quickly and with lower initial costs.
- E. Incorrect.
While integration with AWS managed services is important, both serverless and provisioned solutions can integrate with services like Amazon S3 or DynamoDB, so this is not a deciding factor.