
AWS EMR: Serverless vs EC2 vs EKS — When to Use Each
EMR Serverless vs EC2 vs EKS for Spark. July 2026 refresh — Spot limits, cold start, Iceberg on all three, and a decision matrix for intermittent vs 24/7 jobs.
Blog category

EMR Serverless vs EC2 vs EKS for Spark. July 2026 refresh — Spot limits, cold start, Iceberg on all three, and a decision matrix for intermittent vs 24/7 jobs.

Glue 5.1 (Spark 3.5.6, Iceberg 1.10.0) for lakehouse ETL. July 2026 refresh — Iceberg v2 vs v3, Lake Formation writes, compaction, Athena compatibility.

Glue does EL; dbt does T. July 2026 refresh — Glue 5.1 DPU economics, dbt-athena/Redshift, and the combined stack most teams converge on.

Athena still bills ~$5/TB scanned (or Capacity Reservations by DPU). July 2026 refresh — partitions, Parquet/Iceberg, workgroups, and when provisioned capacity wins.

Embedding QuickSight dashboards in your SaaS product gives every customer analytics without building a BI layer from scratch. Here is the complete implementation guide — embedding types, authentication, row-level security, and cost.

Managed Flink for stateful streaming. July 2026 refresh — KDA name EOL, KPU + orchestration KPU, per-second billing, Flink vs Lambda decision matrix.

Glue ETL + Athena SQL without servers — still the default lake path. July 2026: Glue 5.x jobs, Parquet partitions, ~$0.44/DPU-hour + $5/TB scan math, pipeline checklist.

RDS, Aurora, and self-managed Postgres each have a cost breakeven point. This guide covers total cost of ownership, connection pooling with PgBouncer, indexing strategies, and the edge cases that turn Postgres into a billing surprise.

Atlas vs EC2 vs DocumentDB on AWS — TCO flips with DBA time, not $/vCPU. July 2026: index/cache pressure, hot shards, DocumentDB API-compat caveat.

QuickSight isn't a Looker or Tableau replacement — but for analytics bolted onto the AWS data plane, it's the path that doesn't require a new vendor contract. Data sources, SPICE performance, embedded analytics, row-level security, and dashboards that survive scale.

Fintech on AWS isn't about more services — it's about which services you're allowed to use, how you isolate customer funds, and how fast your audit trail responds. Architecture patterns for payments, fraud detection, regulatory compliance, and the AWS primitives that power modern financial platforms.

July 2026: Amazon DataZone domain/project subscriptions on Glue + Lake Formation — pricing is requests/metadata/compute/tokens (not per-asset monthly), with a subscription pilot checklist.