Skip to content

AWS Data Analytics Specialty (DAS-C01) Study Strategy

Study Approach

Phase 1: Foundation (Weeks 1-2)

  1. Data Collection - Kinesis family (Streams, Firehose, Analytics), IoT Core, DMS
  2. Storage - S3 storage classes, data formats (Parquet, ORC, Avro), partitioning
  3. Glue Fundamentals - Crawlers, Data Catalog, ETL basics
  4. 📖 Kinesis Developer Guide
  5. 📖 Glue Developer Guide

Phase 2: Processing and Analysis (Weeks 3-4)

  1. EMR - Spark, Hive, Presto - cluster configuration and optimization
  2. Glue ETL - DynamicFrame, bookmarks, transforms, job types
  3. Redshift - Distribution styles, sort keys, COPY, WLM, Spectrum
  4. Athena - Cost optimization, federated queries, workgroups
  5. QuickSight - SPICE, data sources, security (RLS, CLS)
  6. 📖 EMR Management Guide
  7. 📖 Redshift Database Developer Guide

Phase 3: Security and Integration (Week 5)

  1. Encryption - KMS, SSE-S3, SSE-KMS, in-transit encryption
  2. Lake Formation - Permissions model, LF-Tags, cross-account sharing
  3. VPC - Endpoints, Enhanced VPC Routing, network isolation
  4. Orchestration - Step Functions, Glue Workflows
  5. 📖 Lake Formation Developer Guide
  6. 📖 KMS Developer Guide

Phase 4: Exam Prep (Week 6)

  1. Practice scenario-based questions
  2. Review service selection decision trees
  3. Focus on integration patterns between services
  4. Take practice exams and review weak areas

Primary Resources

Hands-On Practice

  • Build a streaming pipeline: Kinesis Data Streams - Firehose - S3 - Athena
  • Create a Glue ETL job with crawlers and bookmarks
  • Configure Lake Formation with column-level security
  • Set up Redshift with Spectrum for data lake queries
  • Deploy an EMR cluster with Spark and run a PySpark job

Exam Tactics

Keywords to Watch For

  • "Real-time" + "load to destination" - Kinesis Data Firehose
  • "Real-time" + "custom processing" - Kinesis Data Streams
  • "Serverless ETL" - AWS Glue
  • "Petabyte scale" + "Spark/Hive" - Amazon EMR
  • "Ad-hoc SQL on S3" - Amazon Athena
  • "Complex joins and aggregations" - Amazon Redshift
  • "Full-text search" or "log analytics" - OpenSearch
  • "Business dashboards" - Amazon QuickSight
  • "Fine-grained access control" or "column-level" - Lake Formation
  • "Encrypt" + "audit trail" - SSE-KMS (not SSE-S3)
  • "Keep traffic in VPC" - VPC endpoints + Enhanced VPC Routing

Common Pitfalls

  • Kinesis Data Firehose has a minimum 60-second buffer - it is near real-time, not real-time
  • Athena LIMIT clause does not reduce the amount of data scanned or cost
  • Glue Crawlers discover schema - they do not transform data
  • EMR core nodes store HDFS data - do not use Spot instances for core nodes
  • Redshift COPY command is always preferred over INSERT for bulk loading
  • Lake Formation works alongside IAM, not as a replacement
  • S3 Select works on individual objects, not across a dataset

Time Management

  • 180 minutes for 65 questions - approximately 2.5 minutes per question
  • Flag scenario questions that require analysis and return to them
  • Eliminate clearly wrong answers first (usually 2 are obviously wrong)
  • Look for the "most" correct answer - multiple options may seem valid

Readiness Indicators

  • You can explain when to use each Kinesis service
  • You know S3 format tradeoffs (Parquet vs ORC vs Avro vs CSV)
  • You understand Glue components and their interactions
  • You can design EMR clusters with cost optimization
  • You know Redshift distribution styles and sort key selection
  • You can describe Lake Formation permission model
  • You understand encryption options for each analytics service
  • You score 75%+ consistently on practice exams