February Product Update
← Back to blogUncategorized

February Product Update

It’s been a busy month with lots of updates, let’s dive into them:  Readyset Shallow Cache This month we introduced Readyset Shallow Cache, a new caching layer designed to expand Readyset’s query coverage. Readyset’s deep caching engine has always delivered sub-millisecond performance for operational queries that require ultra-low latency refresh. But many production databases serve a mix of operational and analytical queries from the same system. Shallow Cache complements Deep Cache by accel

Readyset

Readyset

2026-03-06 · 3 min read

It’s been a busy month with lots of updates, let’s dive into them:

Readyset Shallow Cache

This month we introduced Readyset Shallow Cache, a new caching layer designed to expand Readyset’s query coverage. Readyset’s deep caching engine has always delivered sub-millisecond performance for operational queries that require ultra-low latency refresh. But many production databases serve a mix of operational and analytical queries from the same system. Shallow Cache complements Deep Cache by accelerating this broader set of query patterns with near 100% query support and zero ramp-up time.

Deep Cache handles mission-critical queries that require real-time freshness, while Shallow Cache optimizes the rest of your workload. You can read more about Readyset Shallow Cache here. If you’re interested in trying Shallow Cache on your workload, send us a note in our Slack community or reach out to hello@readyset.io.

💬 Product Updates

Readyset Core

  • PostGIS Support (POLYGON + Geometry Improvements): Added support for the PostGIS POLYGON type and improved handling of edge cases like POINT EMPTY, expanding Readyset’s caching support for geospatial queries.
  • Shallow Cache Visibility & Control: Introduced SHOW SHALLOW CACHE ENTRIES and SHOW PROXIED SHALLOW QUERIES to inspect cached entries, proxied queries, and associated performance metadata such as last accessed time, refresh time, and execution duration.
  • Smarter Shallow Cache Workers: Updated refresh workers to spawn on demand and shut down after idle periods, reducing upstream connections and resource consumption.
  • Better Array Support in PostgreSQL: Added support for @>, <@, and ||, improved array equality and type coercion, fixed constant-folding edge cases, and strengthened validation to reject incompatible array inputs earlier in query processing.
  • Improved Replication Stability: Reduced restart-related replicator errors and added handling for replicated data arriving before the base node is open, improving restart reliability.
  • MySQL Functional Index Support: Added support for replicating tables with functional indexes when a usable PRIMARY or UNIQUE key is present, expanding compatibility with modern MySQL schemas.
  • Shallow Cache Upstream Switching: Enabled upstream database changes in shallow-cache-only mode via ALTER READYSET CHANGE UPSTREAM TO '...', eliminating the need to redeploy Readyset.
  • And more!

Check out the full release here.

🆕 New on the Blog & YouTube

/brutal-review and the new 80/20 rule

“Good job” doesn’t make software production-ready. In this post, Michael pulls back the curtain on our internal review system and the 80/20 rule behind it: if LLMs help write the code, even more effort should go into stress-testing it. The focus isn’t on generating more output, it’s on adversarial review, deeper tests, and hardening changes before they ship.

MySQL 9.6: Foreign Key Cascade Operations Finally Hit the Binary Log

For years, MySQL handled foreign key cascades inside InnoDB, leaving them out of the binary log and invisible to replication and CDC pipelines. In this post, Marcelo explains what changed in MySQL 9.6, how it affects binlog correctness, and why it matters for systems like Readyset that consume change events in real time.

Replication Internals: Decoding the MySQL Binary Log - Part 1: Introduction and Data Types

Marcelo kicked off a new deep-dive series on MySQL replication internals, starting with a byte-level walkthrough of the binary log format. In Part 1, he breaks down the core data encodings MySQL uses, from little-endian integers to packed integers and TLV structures, laying the groundwork for manually decoding binlog events. If you’re interested in what goes under the hood of MySQL replication, this series is worth a follow.

Vibe Coding a High-Performance App with Readyset QueryPilot

What happens when you let AI build your application and then point it at QueryPilot? Vini put it to the test. In this post, he vibecodes a CRM with no schema changes, no query rewrites, and no manual optimization. He first runs it against MySQL directly, then puts QueryPilot in the path. QueryPilot automatically identifies and caches the right queries, improving performance for code generated exactly the way modern apps are being built.

🧑‍💻Request Features

Keen to use Readyset but your database or cloud environment is not supported? Reach out to support@readyset.io and let us know.

🤝 Get Involved

We love hearing from you! If you want to get involved with Readyset, you can:

Want to see Readyset in action?

Book a demo and see how Readyset can accelerate your database.

Still scaling the hard way?

Modern applications demand instant performance, even under unpredictable load. Readyset helps you eliminate slow queries, stabilize latency, and scale confidently.

Revolutionize your database performance with Readyset

Serve requests at sub-millisecond latencies with the modern database scaling and query caching system for MySQL and PostgreSQL.

Join our newsletter

Stay updated with the latest news, insights, and developments from Readyset — straight to your inbox.

© 2026 Readyset. All rights reserved.