Data Warehousing on AWS Training Data Warehousing on AWS: Trends, Network And Critical Thinking 4 — Questions and Answers
Question 1: A company wants to share live Redshift data with a business partner's AWS account without copying or moving the data. Which feature enables this?
- Amazon S3 Cross-Account Replication
- Redshift Data Sharing (Correct answer)
- AWS Glue Data Catalog sharing
- Amazon QuickSight Reader accounts
Correct answer: Redshift Data Sharing
Redshift Data Sharing allows producers to share live, transactionally consistent data with consumer accounts without copying data or ETL pipelines.
Question 2: What is the primary benefit of using materialized views in Amazon Redshift for frequently run complex queries?
- They allow queries to run on Redshift Spectrum data without a cluster
- They precompute and store query results, reducing repeated computation overhead (Correct answer)
- They automatically partition the underlying base tables
- They enable real-time row-level streaming ingestion
Correct answer: They precompute and store query results, reducing repeated computation overhead
Materialized views store the precomputed result of complex queries, allowing subsequent requests to read the cached result rather than rerunning the full computation.
Question 3: A Redshift cluster's storage is 85% full but compute utilization is low. Using RA3 nodes, what is the most cost-efficient way to add storage capacity?
- Resize the cluster to add more RA3 nodes to increase compute and managed storage
- Migrate data to S3 and query it exclusively through Redshift Spectrum
- Increase the RA3 managed storage quota independently without changing node count (Correct answer)
- Switch to DS2 nodes which have higher local storage density
Correct answer: Increase the RA3 managed storage quota independently without changing node count
RA3 nodes use Redshift Managed Storage (RMS), which can scale storage capacity independently from compute, avoiding the cost of adding unnecessary compute nodes.
Question 4: Which trend is driving the adoption of the 'data mesh' architectural pattern in large enterprises using AWS?
- Centralized data teams managing all data assets for consistent governance
- Domain-oriented, decentralized data ownership with federated governance at scale (Correct answer)
- Replacing all data warehouses with streaming data pipelines
- Moving all analytics workloads to edge computing locations
Correct answer: Domain-oriented, decentralized data ownership with federated governance at scale
Data mesh decentralizes data ownership to domain teams while enforcing federated governance, addressing the scalability bottlenecks of centralized data platform teams.
Question 5: A data engineer must load 500GB of compressed Parquet files from S3 into Redshift as quickly as possible. Which COPY command optimization is most impactful?
- Use a single large Parquet file to reduce COPY command overhead
- Split files into sizes matching the number of Redshift cluster slices for parallel loading (Correct answer)
- Enable COPY with the STATUPDATE OFF option to skip statistics
- Use the INSERT INTO SELECT pattern instead of COPY
Correct answer: Split files into sizes matching the number of Redshift cluster slices for parallel loading
Splitting source files so each Redshift slice processes approximately one file enables fully parallel loading, maximizing ingestion throughput.
Question 6: A team is evaluating whether to use Amazon Redshift or Amazon DynamoDB for a new workload. Which critical thinking factor most strongly favors Redshift?
- The workload requires single-digit millisecond key-value lookups at high throughput
- The workload involves complex multi-table JOINs and aggregations across billions of historical records (Correct answer)
- The workload requires flexible, schema-less document storage
- The workload needs automatic multi-region active-active replication
Correct answer: The workload involves complex multi-table JOINs and aggregations across billions of historical records
Redshift excels at complex analytical queries with JOINs and aggregations over large historical datasets, while DynamoDB is optimized for simple, high-throughput key-value access patterns.
Question 7: Which AWS service complements Amazon Redshift by providing a serverless, pay-per-query engine for ad-hoc analytics directly on S3 data without ETL?
- Amazon EMR
- Amazon Athena (Correct answer)
- AWS Glue Studio
- Amazon Kinesis Data Analytics
Correct answer: Amazon Athena
Amazon Athena uses Presto/Trino to run SQL directly on S3 data with no infrastructure to manage and charges only for data scanned per query.
Data Warehousing on AWS Training Data Warehousing on AWS: Trends, Network And Critical Thinking 4
A company wants to share live Redshift data with a business partner's AWS account without copying or moving the data.
Which feature enables this?