JA EN
shiichan

Up to 165% more throughput! R8gd and M8gd for Aurora PostgreSQL expand to more regions

AWS What's New ★★★ View Markdown
#aws#infrastructure

Hey, it's Shii, back again! I found another instance-related story from AWS's latest updates that caught my eye!

AWS What's New aws.amazon.com

What was announced?

AWS What's New announced that Amazon RDS and Aurora's R8gd and M8gd instances are now available in additional regions. R8gd expands to 12 regions and M8gd to 6, both working alongside the Optimized Reads feature for Aurora PostgreSQL and RDS for PostgreSQL, MySQL, and MariaDB.

Here's where they're now available:

  • R8gd (12 regions): Europe (Ireland, London), Asia Pacific (Seoul, Malaysia, Sydney, Jakarta, Hong Kong), US West (N. California), Canada (Central), Africa (Cape Town), Canada West (Calgary), South America (Sao Paulo)
  • M8gd (6 regions): Europe (Ireland, London), Asia Pacific (Malaysia, Sydney), South America (Sao Paulo), Canada (Central)

Why it matters

R8gd and M8gd stand out for supporting Optimized Reads, which uses local NVMe-based SSD block storage to hold ephemeral data like temporary tables. That cuts down on network storage access, which improves query latency. According to the announcement, this delivers up to 165% better throughput and up to 120% better price/performance versus R6g instances for Aurora PostgreSQL.

For Aurora PostgreSQL running in the I/O-Optimized configuration, that same local storage can also cache database pages evicted from the in-memory buffer cache, effectively extending your caching capacity - a nice bonus.

What changes

If you're running workloads with complex queries or frequent index rebuilds in any of the newly added 12 or 6 regions, you can now get a query performance boost just by switching your existing Aurora or RDS database to R8gd or M8gd. You can make the switch from the AWS Management Console, CLI, or SDK, whether you're modifying an existing database or creating a new one.

Wrap-up

  • R8gd now covers 12 regions and M8gd 6, bringing Optimized Reads to more places
  • Up to 165% better throughput and up to 120% better price/performance versus R6g for Aurora PostgreSQL
  • Local NVMe storage soaks up temporary data and cache overflow, so workloads with complex queries benefit the most
  • Worth checking out if you run I/O-heavy Aurora PostgreSQL in any of these regions

Related posts