I Passed AWS DEA-C01 in 6 Weeks: Here's My Exact Study Plan

Why I Almost Skipped This Cert (And Why I'm Glad I Didn't)

When AWS announced the Data Engineer Associate exam, I honestly rolled my eyes. Another associate cert? I already had the SAA-C03 and figured data engineering knowledge would just come from doing the job.

I was wrong. The DEA-C01 forced me to fill gaps I didn't know I had. I was building data pipelines at work but had never properly understood Glue's crawlers, Lake Formation's permission model, or why you'd pick Kinesis Data Streams over Kinesis Data Firehose in specific scenarios.

Six weeks later, I walked out of the testing center with a pass. Here's exactly how I did it.

Understanding the DEA-C01 Exam

First, let's get the logistics straight. The AWS Data Engineer Associate (DEA-C01) exam is relatively new, launched in early 2024. It's designed for people who build, maintain, and optimize data pipelines on AWS.

Exam Quick Facts

• 65 questions (50 scored + 15 unscored experimental)
• 170 minutes — plenty of time, honestly
• $150 USD
• Passing score: 720/1000
• Valid for 3 years

The Four Domains

Domain 1 is the biggest chunk, and it's where Glue, Kinesis, and Lambda dominate. Domain 3 caught me off guard — it's heavy on monitoring, troubleshooting, and optimization, which feels like it should be simple but the questions are tricky.

Week 1: Foundations — S3, Glue, and the Data Lake Mindset

Don't jump straight into advanced services. Even if you've used S3 a thousand times, the DEA-C01 tests nuances you've probably never thought about.

S3 Deep Dive (Yes, Really)

S3 sounds basic, but the exam gets into:

• S3 partitioning strategies — How partition keys affect query performance in Athena
• S3 event notifications — Triggering Lambda or SQS when new data arrives
• S3 storage classes for data lakes — When to use Intelligent Tiering vs. lifecycle rules
• S3 Select and Glacier Select — Querying data in-place without full retrieval

Spend 2-3 days just on S3. Create a small data lake, set up partitions, query with Athena. This foundation pays off massively later.

AWS Glue: The Heart of the Exam

If I had to pick one service that dominates the DEA-C01, it's AWS Glue. You need to know:

• Glue Crawlers — how they infer schemas, when they fail
• Glue ETL jobs — Spark-based and Python Shell
• Glue Data Catalog — the central metadata repository
• Glue Bookmarks — for incremental processing (this comes up a LOT)
• Glue Schema Registry — for Avro/JSON schema evolution

Week 2: Streaming Data — Kinesis, MSK, and Real-Time Processing

This is where the exam separates "I've used AWS" from "I understand data engineering." Streaming data questions are consistently the hardest on the exam because there are multiple valid approaches and you need to pick the best one.

The Kinesis Family

Know the differences cold:

• Kinesis Data Streams (KDS) — Custom processing, you manage consumers. Use when you need custom logic or multiple consumers reading the same stream.
• Kinesis Data Firehose — Managed delivery to S3, Redshift, OpenSearch, or HTTP. Use when you just need to land data somewhere with minimal transformation.
• Kinesis Data Analytics — SQL or Apache Flink on streaming data. Use for real-time analytics, windowing, anomaly detection.

The classic exam trap: a question describes a scenario where data needs to go to S3 and Redshift simultaneously. Many people pick KDS because "multiple destinations." But the best answer is often Firehose with a delivery stream to each destination, because it's simpler and fully managed.

Amazon MSK vs. Kinesis

Amazon Managed Streaming for Apache Kafka (MSK) shows up in 5-8 questions. The key decision point: MSK is for when you already have Kafka expertise or existing Kafka producers/consumers. For new greenfield projects, Kinesis is usually the preferred AWS-native answer.

Week 3: Data Storage — Redshift, DynamoDB, and RDS

Week 3 is about understanding where data lives after processing. The exam tests your ability to pick the right storage technology for different workloads.

Redshift: The Data Warehouse Star

Redshift questions are everywhere on this exam. Key concepts:

• Distribution styles — KEY, EVEN, ALL. Know when to use each.
• Sort keys — Compound vs. interleaved. The exam loves testing this.
• Redshift Spectrum — Query S3 data directly from Redshift without loading it.
• Concurrency Scaling — Auto-scales read queries. Know the billing model.
• Redshift Serverless — Newer option, increasingly the recommended answer for variable workloads.

DynamoDB for Data Engineers

DynamoDB appears in maybe 4-6 questions, usually around:

• DynamoDB Streams → Lambda → downstream processing
• When to choose DynamoDB vs. RDS for operational data
• Global tables for multi-region data replication

If you've taken the DVA-C02, you'll find the DynamoDB content here much lighter. The DEA-C01 cares more about DynamoDB as a data source than about table design.

Week 4: Orchestration — Step Functions, MWAA, and EventBridge

Data pipelines don't run in isolation. You need to orchestrate them — trigger jobs, handle failures, manage dependencies. This is Domain 3 territory, and it's trickier than it sounds.

Step Functions: The Orchestrator

AWS Step Functions is the go-to orchestration answer for most exam questions. Know:

• Standard vs Express workflows — Standard for long-running (up to 1 year), Express for high-volume short tasks
• Error handling — Retry policies, Catch blocks, Fallback states
• Integration patterns — How Step Functions triggers Glue, Lambda, ECS tasks
• Parallel and Map states — For processing multiple items concurrently

MWAA (Managed Apache Airflow)

MWAA questions are less frequent but they appear. The main exam scenario: you have an existing Airflow-based pipeline on-premises and need to migrate to AWS. MWAA is the answer. For new pipelines, Step Functions is usually preferred.

EventBridge

EventBridge connects everything. Schedule Glue jobs, respond to S3 events, trigger Step Functions. Know the difference between scheduled rules and event-pattern rules. The exam will test both.

Week 5: Security, Governance, and Lake Formation

Domain 4 is "only" 18% of the exam, but the questions are dense. Security questions require precise knowledge — close enough isn't good enough.

Lake Formation: The Governance Layer

This is probably the most under-studied topic on the DEA-C01, and it shows up more than people expect:

• Fine-grained access control — Column-level and row-level security on data lake tables
• Data permissions model — How Lake Formation permissions interact with IAM
• Cross-account sharing — Sharing data catalog tables across AWS accounts
• Governed tables — ACID transactions in S3 (think Delta Lake but AWS-native)

Encryption and Access Control

• KMS encryption at rest for S3, Redshift, DynamoDB — know the key management options
• VPC endpoints for private access to data services
• IAM roles vs. resource-based policies for cross-service access
• CloudTrail for auditing data access patterns

For the broader AWS security picture, our AWS security certification path covers the full roadmap.

Week 6: Practice Exams and Gap Analysis

This is the most important week. Everything before this was building knowledge. Week 6 is where you convert knowledge into exam-passing ability.

My Practice Exam Schedule

• Day 1: Full practice exam #1 (timed). Score it honestly. Identify weak domains.
• Day 2: Deep review of every wrong answer from exam #1. Study the weak topics.
• Day 3: Targeted practice — 30 questions focused on your weakest domain.
• Day 4: Full practice exam #2. You should see improvement.
• Day 5: Review exam #2 wrong answers. Focus on patterns — are you consistently missing the same type of question?
• Day 6: Light review. Read through your notes. Don't cram new material.
• Day 7: Exam day. You're ready.

The Mistakes I Made (So You Don't Have To)

Real talk — here are the things I got wrong during my prep:

Mistake 1: Ignoring Glue Bookmarks

I figured bookmarks were a minor feature. Nope. At least 3-4 questions on my exam tested bookmark behavior — what happens when a bookmark job fails, how to reset bookmarks, when bookmarks don't work (hint: they don't work well with certain file formats).

Mistake 2: Skimming Lake Formation

I spent maybe 2 hours on Lake Formation and it cost me. The permissions model is its own universe, and the exam tests specific scenarios like "User A needs to query a table but only see rows where region=US." That's Lake Formation row-level security.

Mistake 3: Not Doing Enough Hands-On

Reading about Redshift distribution keys is one thing. Actually creating a table with KEY distribution, loading data, and running EXPLAIN to see query plans is completely different. The exam asks practical questions that you can only answer from experience.

Resources I Used

The Winners

• ExamCert DEA-C01 practice tests — Free, solid explanations, and the difficulty level matched the real exam closely
• AWS Documentation — Specifically the "Best Practices" pages for Glue, Kinesis, and Redshift
• AWS Skill Builder — The free DEA-C01 exam prep course is actually decent
• Hands-on labs — Built 3 end-to-end pipelines using free tier. Worth every hour.

The Losers

• Generic "AWS data engineering" YouTube videos from 2023 — outdated, don't cover DEA-C01 specific topics
• Udemy courses that just read the documentation to you — save your money

Already passed? Consider the full AWS data engineer career path or expand to the AWS Machine Learning Engineer exam next.

If you're still building your AWS foundation, start with our CLF-C02 Cloud Practitioner guide and work up. And for the best study techniques, check our active recall guide.

Comparing AWS data certs? Our AWS certification for beginners article maps out the full landscape.

Frequently Asked Questions