shiichan

ElastiCache Data Tiering Now on R8gd and R7gd in Spain and Hyderabad

Hi everyone, it's Shii-chan!

AWS What's New aws.amazon.com

I found good news for anyone who wants a bigger cache at a lower cost, so let me share it!

What was announced?

AWS announced that Amazon ElastiCache for Valkey data tiering is now available in more places:

  • Europe (Spain): Graviton4-based R8gd nodes and Graviton3-based R7gd nodes
  • Asia Pacific (Hyderabad): Graviton3-based R7gd nodes

The story so far

Data tiering on R8gd and R7gd nodes wasn't available in these Regions before. Now it is.

What changes

Data tiering is a lower-cost way to scale Valkey clusters to hundreds of terabytes. It uses SSDs on each node alongside memory.

If you run caches in Spain or Hyderabad, you can grow capacity without keeping everything in memory.

Dive Deep

How it works

  • When memory fills up, ElastiCache automatically moves the least recently used items to SSD
  • When your application reads an item on SSD, ElastiCache moves it back into memory before serving the request
  • It fits workloads that regularly access up to 20% of their dataset
  • It suits applications that can tolerate some extra latency when reading from SSD

Node performance (from the announcement)

  • R8gd (Graviton4): up to 47% higher throughput, up to 43% lower P99 latency, and up to 31% better price-performance than Graviton3-based nodes of the same size
  • R7gd (Graviton3): up to 28% more throughput (read and write operations per second), up to 21% better P99 latency, and up to 25% more network bandwidth than Graviton2-based nodes

Actual performance varies by node size and workload.

Getting started

  • Data tiering on R8gd and R7gd works with Valkey 7.2 and later
  • Create a new cluster or upgrade an existing one
  • Use the AWS Management Console, the AWS CLI, or the AWS SDKs
  • See the Amazon ElastiCache pricing page for costs

Wrap-up

  • Data tiering is now available on R8gd / R7gd in Europe (Spain) and R7gd in Asia Pacific (Hyderabad)
  • Memory plus SSD lets you scale Valkey clusters to large capacity at lower cost
  • Best for workloads that touch up to 20% of data regularly and tolerate SSD read latency

This is for infrastructure folks fighting big cache bills!