AWS Cloud Architect & AI Expert
On a composite ingest workload (~8k ordered TPS, 1 KB payloads), staying on SQS FIFO without high-throughput mode capped effective throughput near 300 TPS/API and modeled queue backlog cost near $95/mo before ops time. Enabling high-throughput FIFO or switching to Kinesis on-demand raised the ceiling without changing the consumer code.
On a composite B2B order API (~2.4k RPS peak, 24 ECS tasks on Aurora), running the scaling ladder in order (RDS Proxy, read-replica routing, ElastiCache, then SQS offload) moved modeled p99 from 1,840 ms to 72 ms before any microservice split.
A composite fintech API (~8ms p99 to metro users) spent ~$22k/mo on hybrid infrastructure around an Outposts rack before moving read-heavy paths to a Chicago Local Zone and keeping Outposts only for card-PIN HSM proximity. p99 held at 9ms and hybrid infra dropped ~38%. This post covers placement and connectivity, not EC2-vs-Lambda unit economics.
On a composite mid-market OTT platform (~180k concurrent live peak, ~12 TB/month VOD catalog), routing Saturday-night live through MediaConvert batch queues added ~$4,200/mo in idle transcode capacity before switching live to IVS and keeping MediaConvert for catalog only. CloudFront egress still dominates steady-state, so model it before you buy encoders.
Copying curated Parquet into every consumer account is how data platforms drown in storage cost and permission sprawl. On Feb 11, 2026 AWS shipped Lake Formation cross-account version 5, which adds wildcard RAM shares for hundreds of thousands of tables. A composite 12-account platform cut duplicate curated copies from 3 to 0 and dropped cross-account access tickets from ~11/month to ~3 by standardizing LF-Tags + resource links.
AWS Blocks (public preview, June 2026): 18+ open-source TypeScript building blocks that run locally without an AWS account and deploy with zero code changes.
Bloom filters shave 90% of negative lookups; HyperLogLog estimates cardinality without storing every user ID. Redis modules on ElastiCache for abuse detection and feed deduplication.
Why Aurora PostgreSQL loves B-tree indexes on OLTP but DynamoDB feels like an LSM, and how cost-based optimization surprises you when statistics go stale on RDS.
CAP explains why your global DynamoDB table shows stale inventory and why Aurora Global reads lag 80 ms behind the writer. This guide maps partition tolerance, consistency, and availability trade-offs to concrete AWS controls.
You rarely implement Raft on EC2, because you buy it in Aurora, DynamoDB, and EKS etcd. This guide explains quorum math so you can trust managed services and avoid rolling your own coordinator.
Two goroutines updating adjacent counters can saturate the memory bus on a c7g.4xlarge. This guide covers memory barriers, cache lines, and false sharing, and why placement groups do not fix application-level contention.
Whole-item last-write-wins is how Global Tables lose cart merges. When to use counters, OR-Sets, and conflict-free merges vs when to keep a single Aurora writer.
Our articles share what we know. Our consulting engagements apply that knowledge to your specific AWS environment.
