starter.wiki

AWS managed application cache

Amazon ElastiCache

An AWS managed cache service offering Serverless and node-based deployments with Valkey, Redis OSS, and Memcached engines.

Visit official site

Editorial verdict

Choose Amazon ElastiCache when the application is AWS-native and needs a managed disposable cache with either Serverless elasticity or explicit node capacity. Select the engine and deployment independently: default to Valkey for new compatible workloads only after client and command validation, retain Redis OSS when its exact compatibility is required, and choose Memcached for the simpler multithreaded object-cache model.123

Best for

  • AWS applications that need a low-latency shared cache inside an intentional VPC and Region design
  • Teams choosing Serverless to avoid capacity planning or node-based clusters for explicit topology and configuration control
  • Disposable workloads whose required Valkey, Redis OSS, or Memcached commands and clients match the selected ElastiCache deployment
123

Not ideal for

  • Authoritative or irreplaceable application data that cannot be rebuilt after eviction or cache loss
  • Globally distributed applications expecting one ElastiCache cache to provide transparent multi-region access
  • Clients or workloads that require restricted commands, non-TLS Serverless access, modifiable Serverless parameters, or unsupported engine behavior
12345
Main trade-off

ElastiCache reduces cache infrastructure operations and integrates with AWS networking, but the application accepts Region and VPC locality, engine and command compatibility, Serverless ECPU and GB-hour economics or node-hour capacity planning, eviction behavior, and cross-AZ transfer costs.123457

Product boundary

Whether an AWS-native application cache should use ElastiCache Serverless or node-based capacity, and which supported engine fits the cache semantics.

ElastiCache is an in-memory application cache, not the system of record. Serverless abstracts cluster capacity behind a TLS endpoint and fixed cluster-mode configuration; node-based deployments expose node type, node count, placement, shards, replicas, and parameter choices. Engine compatibility is managed and version-specific rather than identical to every self-operated Valkey, Redis OSS, or Memcached deployment.123

For: AWS application teams that need a disposable shared cache and can intentionally choose Valkey, Redis OSS, or Memcached plus Serverless or node-based deployment

  • The cache requires a command, module, client mode, or parameter that Serverless or the selected engine does not support.
  • Measured stored GB-hours and ECPUs make node-based capacity cheaper, or measured node utilization makes Serverless more appropriate.
  • VPC endpoint placement, cross-AZ traffic, Region availability, or private connectivity does not match the application topology.
  • The data cannot safely tolerate TTL expiry, eviction, cache loss, or eventually consistent replica reads.

Why teams consider Amazon ElastiCache

  • Two deployment modelsServerless automatically manages cache capacity, while node-based clusters expose node, shard, replica, Availability Zone, and parameter choices.123
  • Three engine routesElastiCache supports Valkey, Redis OSS, and Memcached with materially different data structures, threading, replication, persistence, and clustering behavior.123
  • AWS availability integrationServerless stores data redundantly across multiple Availability Zones, while node-based clusters let teams choose single-AZ or multi-AZ topology.123
  • Explicit cost dimensionsServerless meters stored GB-hours and ECPUs; node-based clusters meter provisioned node-hours with optional reservations.1237

Pricing

ElastiCache Serverless bills cached data in GB-hours plus requests in ECPUs, with a lower 100 MB minimum data-storage meter for Valkey and a 1 GB minimum for Redis OSS and Memcached. Node-based clusters bill by engine, node type, node count, and node-hours; backups, cross-AZ transfer, and other AWS dimensions can add cost.7

Elastic or uncertain cache demand

ElastiCache Serverless

Billed by average data stored in GB-hours and ECPUs based on compute time or transferred data, whichever drives the request; rates vary by engine and AWS Region.7

Explicit capacity and topology

Node-based cluster

Billed by provisioned node-hours according to engine, node type, node count, and Region, with one-year or three-year reserved-node options for eligible steady workloads.7

Serverless request meter
ECPUs reflect vCPU time and transferred data; simple reads and writes begin at 1 ECPU per KB transferred7
Serverless storage floor
100 MB for Valkey; 1 GB for Redis OSS and Memcached under current pricing7
Data transfer boundary
No EC2-to-ElastiCache charge in the same AZ; standard EC2 cross-AZ charges can apply when endpoints or nodes are in another AZ7
Pricing checked View official pricing

Amazon ElastiCache vs alternatives

Upstash Redis

Choose when
Choose it when HTTP REST access, request-oriented serverless pricing, and supported Redis-compatible semantics matter more than AWS VPC-native deployment.
Compared with Amazon ElastiCache
The application accepts Upstash's command compatibility, eventual consistency, storage, bandwidth, region, and redundancy boundaries.8

Redis Cloud

Choose when
Choose it when managed Redis breadth, cross-cloud availability, Redis Flex, or Redis-specific enterprise topology matters more than AWS-native ElastiCache integration.
Compared with Amazon ElastiCache
The service has its own capacity, persistence, topology, provider, region, and commercial plan boundaries.9

Resources and sources

Official product, policy, and pricing

  • What is Amazon ElastiCache
    Open
  • ElastiCache deployment options
    Open
  • Comparing ElastiCache engines
    Open
  • ElastiCache supported and restricted commands
    Open
  • ElastiCache Valkey and Redis OSS limits
    Open
  • ElastiCache service quotas
    Open
  • Amazon ElastiCache pricing
    Open
  1. 1
    What is Amazon ElastiCache

    Amazon Web Services · Accessed Official

  2. 2
    ElastiCache deployment options

    Amazon Web Services · Accessed Official

  3. 3
    Comparing ElastiCache engines

    Amazon Web Services · Accessed Official

  4. 4
    ElastiCache supported and restricted commands

    Amazon Web Services · Accessed Official

  5. 5
    ElastiCache Valkey and Redis OSS limits

    Amazon Web Services · Accessed Official

  6. 6
    ElastiCache service quotas

    Amazon Web Services · Accessed Official

  7. 7
    Amazon ElastiCache pricing

    Amazon Web Services · Accessed Official

  8. 8
    Upstash Redis getting started

    Upstash · Accessed Official

  9. 9
    Redis Cloud documentation

    Redis · Accessed Official