redis

ElastiCache bills node-hours for Redis/Valkey clusters and ECPUs for Serverless. A cache.r7g.large with two replicas is 3× the primary node cost before data transfer. Serverless wins below ~$800/mo steady cache spend; provisioned nodes win above it.

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.

Valkey on ElastiCache: about 20% node / 33% serverless cheaper than Redis OSS engines. July 2026 — sessions, rate limits, locks, migration checklist.

SQS, MSK Kafka, and Redis queues are not interchangeable. Each has different cost models, ordering guarantees, and failure modes. This guide covers when to use each, how to autoscale workers on queue depth, and how to build idempotent consumers.

Redis is fast — until your application retries on every cache miss and the Redis bill starts looking like the database bill. ElastiCache patterns, data structures, cluster modes, eviction policies, and the production patterns that actually reduce database load.

ElastiCache loses your AI chatbot's session memory at every node replacement. MemoryDB doesn't. A decision framework for when to pick MemoryDB over ElastiCache, OpenSearch Serverless, and S3 Vectors for AI workloads — with the latency math and the failure mode that forces the switch.