Domain 4 of 4 · Chapter 4 of 4

Modernization and enhancement opportunities

Unlock the complete study guide + 1,040 practice questions across 16 full exams.

Bundled into the existing AWS Certified Solutions Architect - Professional premium course — no separate purchase.

Included in this chapter:

  • Spotting modernization opportunities in an existing workload
  • Managed-service adoption: deleting operational debt
  • Decoupling: event-driven and serverless modernization
  • Data and analytics modernization on AWS
  • Mainframe modernization and exam pattern recognition

Self-managed engine to AWS managed equivalent

Self-managed on EC2AWS managed equivalentCompatibility / protocolWhat AWS now operates
Apache KafkaAmazon MSKOpen-source Apache Kafka API (apps unchanged)Brokers, Kafka metadata (ZooKeeper/KRaft), broker auto-recovery; MSK Serverless / Connect / Replicator
Redis / MemcachedAmazon ElastiCacheRedis and Memcached wire protocolsNode provisioning, patching, failover, backups, scaling
ElasticsearchAmazon OpenSearch ServiceOpenSearch + legacy Elasticsearch OSS up to 7.10Cluster provisioning, failed-node replacement, snapshots; OpenSearch Serverless
Hadoop / SparkAmazon EMROpen-source Hadoop, Spark, Hive, Presto, etc.Cluster lifecycle, scaling, Spot integration; EMR Serverless / EMR on EKS
COBOL/PL-I mainframeAWS Mainframe ModernizationRefactor to Java OR recompile same languageManaged runtime + tooling; refactor (AWS Transform for mainframe) vs replatform (Rocket/Micro Focus)

Decision tree

What is the dominant modernization signal? Self-managed OSS engine Coupled / scale-for-peak Data / mainframe Which engine on EC2? Kafka / cache / search / Hadoop Amazon MSK Kafka ElastiCache Redis / Memcached OpenSearch Service Elasticsearch Amazon EMR Hadoop / Spark Sync glue or stateful long-running service? Event-driven glue SQS / SNS / EventBridge + Lambda (scale to zero) Multi-step workflow Step Functions Standard = exactly-once Container / long-running Fargate no EC2 to manage Data layer or legacy mainframe? One DB / analytics Purpose-built DB + S3 data lake (Lake Formation) COBOL mainframe Mainframe Modernization refactor (Java) vs replatform Always: drive priorities from a Well-Architected review; modernize incrementally (strangler-fig), not big-bang
Cheat sheet

Cheat sheet

  • Migration is the start of modernization, not the end
  • Well-Architected reviews surface the modernization backlog
  • Move self-managed Kafka on EC2 to Amazon MSK
  • MSK auto-recovers a failed broker and keeps the same IP
  • MSK has Serverless, Connect, and Replicator variants
  • Move a self-managed cache fleet to Amazon ElastiCache
  • Move self-managed Elasticsearch to Amazon OpenSearch Service
  • Move self-managed Hadoop/Spark to Amazon EMR
  • Managed-service adoption preserves the protocol and deletes the toil
  • Lambda's asynchronous event payload limit is 1 MB
  • Async Lambda retries function errors twice; route failures off
  • An SQS dead-letter queue must match its source queue type
  • Front decomposed microservices with Amazon API Gateway
  • Build the analytics estate as an S3 data lake
  • Lake Formation centralizes fine-grained, cross-service governance
  • Mainframe Modernization: refactor to Java vs replatform in place
  • Don't modernize a workload marked Retire

Unlock with Premium — includes all practice exams and the complete study guide.

References

  1. AWS Well-Architected Framework
  2. What is Amazon MSK?
  3. What is Amazon ElastiCache?
  4. What is Amazon OpenSearch Service?
  5. What is Amazon EMR?
  6. What is Amazon Simple Queue Service?
  7. What is Amazon SNS?
  8. What is Amazon EventBridge?
  9. What is AWS Step Functions?
  10. What is AWS Lambda?
  11. AWS Fargate for Amazon ECS
  12. What is Amazon DynamoDB?
  13. Using Aurora Serverless v2
  14. What is Amazon S3?
  15. What is AWS Glue?
  16. What is Amazon Athena?
  17. Amazon Redshift management overview
  18. What is AWS Lake Formation?
  19. Decomposing monoliths: the strangler fig pattern
  20. What is Amazon API Gateway?
  21. Asynchronous invocation for AWS Lambda
  22. Amazon SQS dead-letter queues
  23. Lake Formation tag-based access control (LF-Tags)
  24. What is AWS Mainframe Modernization?