MySQL benefits from a large and technically rigorous community—people who run MySQL at scale, test it in diverse environments, and surface issues (and ideas) that make the product better. Over the last year, we’ve been listening to community feedback and, since January 2026, sharing our updated community engagement approach (see: A New Era of MySQL Community Engagement). In this post, we describe how we’re improving community conversations and collecting feedback on what’s working, what’s unclear, and where we can improve the collaboration loop. 

This renewed openness and pace of development will succeed with thoughtful input and feedback from users and contributors. The feedback, ideas, and experiences shared in this community continue to shape our direction and strengthen the impact of our work. We are deeply committed to maintaining an open, transparent dialogue as we evolve and improve MySQL together with a clear vision for 2026 and beyond and a community focused three-pronged strategy. First, we will deliver innovation by introducing developer-focused features into the MySQL Community Edition. Second, we aim to extend and enrich the ecosystem and adoption. Finally, we are increasing transparency and encouraging broader community participation, ensuring that more voices help shape MySQL’s evolution.

This post has three parts: first, a few feature highlights in MySQL 9.7.0 Community Edition, second, how you can provide feedback on Early Access (EA) Release builds, and third, publication of Worklogs

What’s new in MySQL 9.7.0 Community Edition: feature highlights

Below are highlights in MySQL 9.7.0 Community Edition. The release notes remain the authoritative source for complete details and final behavior.

Flow-control monitoring for MySQL InnoDB Cluster (Group Replication)

Flow control protects a cluster by throttling the primary when secondaries fall behind. MySQL 9.7.0 adds better visibility into when flow control triggers and how much throttling it causes, including metrics such as: 

  • number of sessions currently being throttled 
  • number of throttling events 
  • total throttle time 
  • timestamp of the last throttled transaction

This helps DBAs understand whether a cluster is under stress, tune flow control settings, and achieve smoother throughput. 

Multi-threaded applier: extended applier statistics

MySQL 9.7.0 enhances replication observability by adding new Performance Schema tables that gather statistics across the replication pipeline: 

This enables better diagnosis of replication lag, improved throughput visibility (including bytes/sec and queued/ongoing work), and clearer insight into applier behavior. 

Telemetry (OpenTelemetry / OTLP)

Telemetry support was introduced earlier in the MySQL 9.x series, and MySQL 9.7.0 continues that direction by making it easier to integrate MySQL into modern observability stacks via the OpenTelemetry Protocol (OTLP). Highlights include: 

  • OTLP-based logs, metrics, and traces exported to an OpenTelemetry collector 
  • component-based architecture (install/configure logging/tracing/metrics independently) 
  • direct-to-collector approach (no additional log agents required) 
  • connector support (e.g., Connector/J and Connector/NET) for end-to-end trace context propagation

Profile-Guided Optimization (PGO)

MySQL 9.7.0 is now built using Profile-Guided Optimization (PGO)—a compiler optimization technique that uses runtime profiling to produce better-optimized binaries, targeting improved performance for real-world workloads. 

MySQL JSON Duality Views

MySQL 9.7.0 introduces JSON Duality Views, allowing developers to work with the same underlying data using both relational SQL and a hierarchical JSON document model—combining relational integrity with JSON-centric agility, while reducing the need for complex ORM mappings or data synchronization. 

Hypergraph Optimizer

MySQL 9.7.0 includes the hypergraph optimizer, designed to improve performance for complex queries by exploring a broader set of join plans (including bushy trees) using a cost-based approach. The hypergraph optimizer supports cost-based selection of hash joins and will take interesting orders into account when choosing the query plan.  This will improve the execution time for a large class of MySQL queries.

Keyring_OCI (OCI Vault integration)

MySQL 9.7.0 integrates with OCI Vault via the component_keyring_oci keyring component, allowing encryption keys to be stored in OCI Vault rather than locally on the MySQL server. 

There are other features ready to integrate into this release, such as Vector support, that we are excited to collaborate with the MySQL Community for more feedback and discussion to ensure we meet the needs to the MySQL Ecosystem before integrating them into a MySQL Community Release. This will be a focus in our ongoing public MySQL Community roadmap and contributions discussions. We anticipate a public discussions on streamlining and improving our MySQL Community Contributions next month, with an initial MySQL Contributors Summit in May.

Early Access (EA) Release builds: pre-GA packages for early feedback

To improve how we build MySQL in collaboration with the community, we are publishing Early Access (EA) Release buildspre-GA packages intended to help the community test features earlier and provide feedback while changes are still easier to incorporate. 

Early Access builds are provided as developer preview packages for feedback and testing. They may change before GA and are not intended for production use

Where to get EA builds: We’ve publish on a dedicated download page.

How to provide feedback (and what’s most valuable)

The most valuable feedback is your experience running EA builds in real environments. In particular, we’d love to hear about: 

  • regressions vs previous versions 
  • upgrade and downgrade issues 
  • performance changes (positive or negative) 
  • replication/HA edge cases (including Group Replication / InnoDB Cluster)
  • feedback on new features?

You can share feedback through any of the following channels: 

When filing a bug, please include: 

  • the exact EA build identifier and platform, eg. 9.7.0-ER, tag as ‘Early Release’
  • steps to reproduce (ideally a minimal repro) 
  • expected vs actual behavior 
  • configuration details and relevant logs/errors

Worklogs: increased transparency (selected publications)

In addition to EA Release builds, we have resumed publishing selected MySQL Worklogs (WL) to increase transparency and invite design feedback:
https://dev.mysql.com/worklog/ 

In open-source development, transparency is key. For MySQL, Worklogs are architectural blueprints used by the development team to design, track, and implement new features or major refactors. Years ago, we regularly published Worklogs after releases—but we haven’t done that consistently for many years. Going forward, we plan to publish selected Worklogs to show intent, share tradeoffs, and get community feedback earlier. 

Worklogs are not a promise of delivery. Published Worklogs will be curated and reviewed to avoid exposing sensitive information (including security-related details that shouldn’t be public before fixes are available). 

How the community can engage (and what feedback is most valuable)

In closing, the most valuable feedback you can give is your experience with EA Release builds: 

  • Test EA Release builds (link coming soon) and share regressions, upgrade issues, performance changes, and replication/HA edge cases 
  • File actionable bug reports with clear reproduction steps and environment details 
  • Review published Worklogs and comment on design tradeoffs or missing cases

Our public MySQL Community Roadmap discussion in February gave us valuable feedback on shaping the roadmap categories for new features. Our next public MySQL Community discussion will be on March 23. Join our next Public MySQL Community Roadmap Discussion Webinar (Edition #2). We will share more updates on the roadmap, ecosystem growth, adoption, and how we can further increase community contributions and transparency over the next few weeks. 


See also

As always, thank you for using MySQL. Your feedback is essential to evolving MySQL together!