Total progress
0%
High priority done
0/0
S3, Glue, Athena, Kinesis, Redshift, security.
Target readiness
Not ready
Ready when 2 practice exams ≥80% and weak topics closed.
Next action
Start Day 1
Auto-calculated from unchecked daily tasks.
Fastest realistic strategy
1. Do not watch passively.
Watch the Udemy lessons at 1.25–1.5x, then immediately answer 10–20 questions on the topic.
Watch the Udemy lessons at 1.25–1.5x, then immediately answer 10–20 questions on the topic.
2. Build reflexes.
For every scenario, identify: source → latency → transformation → target → governance → monitoring.
For every scenario, identify: source → latency → transformation → target → governance → monitoring.
3. Error log is mandatory.
Every wrong answer becomes a rule. Revisit the rule 24h later and during final review.
Every wrong answer becomes a rule. Revisit the rule 24h later and during final review.
Exam focus by official weight
| Domain | Weight | Fast-track implication |
|---|---|---|
| D1 Data Ingestion & Transformation | 34% | Highest priority: Kinesis, Firehose, Glue, Lambda, EventBridge, Step Functions, DMS, formats. |
| D2 Data Store Management | 26% | S3, Redshift, DynamoDB, Glue Catalog, partitions, lifecycle, schema evolution, Iceberg. |
| D3 Data Operations & Support | 22% | Monitoring, troubleshooting, data quality, Athena/Redshift SQL, CloudWatch, CloudTrail, Glue failures. |
| D4 Security & Governance | 18% | IAM, KMS, Lake Formation, Macie, Secrets Manager, audit logs, data sharing, sovereignty. |
Master Tracker
Complete in this order. “High” means exam-heavy or frequently confusing; “Medium” means know service selection and key trade-offs; “Low” means skim unless practice exams expose weakness.
| Done | Priority | Exam domain | Udemy course block / lessons | What to know for the exam | Confidence | Notes |
|---|
Weekly / Daily Schedule
Aggressive 3-week plan. If a day is too heavy, move its review block to the weekend, but keep practice questions from Day 5 onwards.
AWS exam guide ↔ Udemy course mapping
| AWS domain / task | Udemy sections to complete | Fast-track treatment |
|---|---|---|
| D1.1 Ingestion: streaming/batch, APIs, schedulers, events, Lambda from Kinesis, throttling, fan-in/out, replayability. | Analytics: Kinesis Streams, Firehose, MSAF/Flink, MSK; Migration: DMS/DataSync; App Integration: EventBridge, SQS/SNS; Storage: S3 event notifications. | Deep Memorize service-selection rules and failure modes. |
| D1.2 Transform/process: EMR, Glue, Lambda, Redshift, format conversion, cost/performance. | Analytics: Glue, EMR, Athena Spark; Compute: Lambda; Containers: ECS/EKS basics. | Deep Glue vs Lambda vs EMR vs Redshift decisions. |
| D1.3 Orchestration: MWAA, Step Functions, Glue workflows, EventBridge, notifications. | Application Integration: Step Functions, EventBridge, MWAA, SQS/SNS; Analytics: Glue Workflows/Bookmarks. | Deep Trigger, retry, DLQ, state machine, DAG. |
| D1.4 Programming concepts: Lambda tuning, SAM, IaC, CI/CD, Git, distributed computing. | Fundamentals: Git/SQL; Compute: Lambda & SAM; Developer Tools: CloudFormation/CDK/CodePipeline if present. | Selective Concepts only; no language syntax. |
| D2.1 Choose data store: Redshift, RDS, DynamoDB, EMR, Lake Formation, vectors, Iceberg. | Storage; Database; Analytics; GenAI/vector lessons if present. | Deep Data-store selection table. |
| D2.2 Cataloging: Glue Catalog, crawlers, Hive metastore, partitions, connections. | Analytics: Glue, Hive, Glue Catalog; Storage: S3 Tables/Iceberg. | Deep Catalog + partition sync + schema discovery. |
| D2.3 Lifecycle: S3 lifecycle, versioning, DynamoDB TTL, Redshift COPY/UNLOAD. | Storage: S3 lifecycle/versioning; Database: DynamoDB TTL; Database/Analytics: Redshift COPY/UNLOAD. | Deep Cost + retention + legal deletion patterns. |
| D2.4 Models/schema: Redshift, DynamoDB, Lake Formation, DMS/SCT, lineage, partitions/compression. | Fundamentals: modeling/schema evolution; Database: DynamoDB/Redshift; Migration: DMS/SCT; Analytics: Lake Formation. | Deep Star schema vs key-value vs lakehouse. |
| D3 Operations: automate, analyze, monitor, troubleshoot, data quality. | Analytics: Glue Data Quality/DataBrew, Athena, Redshift; Management: CloudWatch/CloudTrail/Config; App Integration. | Practice-heavy Learn via wrong answers and scenarios. |
| D4 Security/Governance: IAM, KMS, Lake Formation, Macie, Secrets Manager, CloudTrail Lake, privacy. | Security, Identity & Compliance; Analytics: Lake Formation; Storage: S3 encryption/access points; Redshift security. | Deep Least privilege + fine-grained governance + audit. |
Practice Exam Tracker
| Exam | Date | Score % | Pass? | Top weak domain | Action |
|---|
Error Log
Do not just write the correct answer. Write the rule you missed.
| Topic | Wrong assumption | Correct rule to memorize | Closed? |
|---|
Cheat Sheet – Domain 1: Data Ingestion & Transformation (34%)
Core decision tree
- Streaming, replay, multiple consumers: Kinesis Data Streams or MSK.
- Streaming to S3/Redshift/OpenSearch with low ops: Kinesis Data Firehose.
- CDC from databases: AWS DMS.
- SaaS → S3/Redshift: AppFlow.
- Files on-prem → S3/EFS/FSx: DataSync or Transfer Family.
Transform choice
- Small/event transform: Lambda.
- Serverless Spark ETL: Glue.
- Big Spark/Hadoop control: EMR.
- SQL transform in warehouse: Redshift.
- Streaming analytics/windowing: Managed Service for Apache Flink.
Kinesis must-know
- Partition key decides shard; bad keys create hot shards.
- Enhanced fan-out gives dedicated throughput per consumer.
- Producers can create duplicates; consumers must be idempotent.
- Data Streams = replayable. Firehose = delivery service, less consumer control.
Orchestration rules
- Event-driven: EventBridge.
- Visual state machine, retries, audit trail: Step Functions.
- Airflow DAGs: MWAA.
- Glue-native ETL chains: Glue Workflows + bookmarks.
- Failure isolation: SQS DLQ / SNS alerts.
Exam traps
- Do not choose Lambda for long-running heavy ETL.
- Do not choose Firehose when you need multiple independent consumers/replay semantics.
- Do not choose EMR when the question says least operational overhead and Glue fits.
- Format conversion to Parquet is often part of the best answer.
One-line memory hooks
- Low ops streaming delivery → Firehose.
- Replay + fan-out → Kinesis Data Streams.
- Batch Spark ETL → Glue.
- Complex DAG → MWAA; simple serverless workflow → Step Functions.
Cheat Sheet – Domain 2: Data Store Management (26%)
Store selection
- S3: durable data lake/object storage.
- Athena: serverless SQL on S3.
- Redshift: high-performance warehouse/BI.
- DynamoDB: low-latency key-value/document at scale.
- RDS/Aurora: relational transactions.
- OpenSearch: search/log analytics.
S3 lake optimization
- Use Parquet/ORC for analytics.
- Partition by high-value filters, not by ultra-high-cardinality fields.
- Compress to reduce scan cost.
- Lifecycle policies for cost and retention.
- Versioning protects against accidental delete/overwrite.
Glue Data Catalog
- Crawlers infer schema and populate tables.
- Partitions must be registered/synchronized.
- Glue Catalog can act as Hive metastore.
- Catalog + Lake Formation enables governed discovery/access.
Redshift must-know
- COPY loads from S3; UNLOAD exports to S3.
- Spectrum queries S3 external tables.
- Materialized views speed repeated queries.
- Federated query accesses live RDS/Aurora data.
- RA3 decouples compute/storage.
Schema/model traps
- Redshift: star/snowflake, sort/dist choices.
- DynamoDB: design by access patterns; avoid scans/joins.
- Schema evolution: Glue Catalog + compatible formats help.
- Iceberg/open table formats matter for lakehouse tables.
Lifecycle hooks
- DynamoDB TTL removes old items.
- S3 lifecycle transitions/expiration control cost/compliance.
- Legal deletion requirement means explicit deletion/expiry policy, not just cheaper storage.
Cheat Sheet – Domain 3: Data Operations & Support (22%)
Monitoring stack
- CloudWatch Logs: application/service logs.
- CloudWatch Metrics/Alarms: operational alerting.
- CloudTrail: API activity and audit.
- CloudTrail Lake: centralized queryable audit events.
- Config: configuration history/compliance.
Glue troubleshooting
- Check IAM role, network/VPC endpoints, source credentials.
- Bookmark issues cause duplicate/missing processing.
- Partition/schema mismatch breaks Athena/Glue queries.
- Skew causes slow Spark jobs; repartition/salt/change key.
Kinesis troubleshooting
- Hot shard → bad partition key.
- Producer throttling → batch, retry, increase shards/on-demand.
- Consumer lag → scale consumers/enhanced fan-out.
- Duplicates → idempotent consumer/dedup key.
Data quality
- Completeness, consistency, validity, accuracy, uniqueness.
- Glue Data Quality rules can fail job or publish metrics.
- DataBrew for visual profiling/cleaning.
- Sampling: random, stratified; skew requires special handling.
Analysis rules
- Athena for ad hoc S3 SQL.
- Redshift for repeated, high-performance analytics.
- QuickSight for visualization.
- Provisioned vs serverless is usually cost predictability/control vs low ops/elasticity.
Exam traps
- Audit question → CloudTrail, not CloudWatch.
- Application logs → CloudWatch Logs.
- Configuration drift → AWS Config.
- Data quality during ETL → Glue Data Quality/DataBrew, not only Athena.
Cheat Sheet – Domain 4: Data Security & Governance (18%)
Access control
- Least privilege always wins.
- IAM roles for AWS services; avoid long-lived credentials.
- Resource policies for S3/cross-account access.
- S3 Access Points simplify access at scale.
- PrivateLink/VPC endpoints for private service access.
Lake Formation
- Fine-grained data lake permissions.
- Works with S3, Athena, Redshift, EMR.
- Column/table/database-level governance.
- Use when IAM/S3 bucket policy is too coarse.
Encryption
- KMS for managed key control and audit.
- SSE-S3: simple default S3 encryption.
- SSE-KMS: stricter control/audit/cross-account considerations.
- TLS/HTTPS for in transit.
- Secrets Manager for credential storage/rotation.
Privacy & governance
- Macie identifies sensitive/PII in S3.
- Mask/anonymize where compliance requires.
- SCP/IAM/S3 policies can restrict disallowed regions.
- Data sovereignty = region/location control + audit.
Audit
- CloudTrail tracks API calls.
- CloudWatch Logs stores app/service logs.
- CloudTrail Lake enables SQL-like audit queries.
- Athena/OpenSearch can analyze large log datasets.
Exam traps
- Column-level data lake permissions → Lake Formation.
- Find PII in S3 → Macie.
- Rotate DB password → Secrets Manager.
- Who changed config? → Config/CloudTrail depending on wording.
Backup, Import / Export & Profiles
Autosave is enabled.
Every checkbox, score, note and error-log entry is saved automatically in this browser under the selected profile.
Every checkbox, score, note and error-log entry is saved automatically in this browser under the selected profile.
Active profile
JSON backup
Export creates a timestamped JSON file with your current profile progress. Import restores values into the selected profile.
No JSON imported/exported yet.
Local versioned snapshots
Create named snapshots before major changes or practice exams. Snapshots stay in this browser and can also be downloaded as JSON.
Reset options
Resources & exam-day checklist
| Resource | Use |
|---|---|
| Official AWS certification page | Exam logistics: 130 minutes, 65 questions, exam cost, languages, official preparation steps. |
| Official AWS DEA-C01 exam guide | Source of truth for domains, weights, task statements, and in-scope services. |
| Udemy course | Main learning path; use at 1.25–1.5x and skip low-yield sections after you pass practice thresholds. |
| Practice exams | Minimum: 3 full exams. Ready when consistently ≥80–85% and no repeated weak domain. |
Book the exam when: you finish all high-priority rows, score ≥80% on two full practice exams, and can explain every error-log entry without notes.
Use the Backup / Profiles page for visible import/export controls, profiles, and versioned snapshots.