At Percona, our mission has always been to help you succeed with open source databases. We do that by listening to the community, understanding the challenges you face, and building the solutions you need. Now, after a comprehensive review of market trends and direct feedback from our customers and the MySQL community, we are excited to announce our plans for two major new initiatives for the MySQL ecosystem: Vector Search & Indexing and a dedicated MySQL Binlog Server.
This isn’t a choice between two different paths; it’s a commitment to two complementary pillars of our strategy. We believe that to be the best partner for you, we must be both a Trusted partner who solves today’s hardest problems and an Innovator who builds for tomorrow’s opportunities. These initiatives deliver on both of those promises.
Vector Support: Embracing the AI-Powered Future
The rise of AI and machine learning is reshaping how applications are built. The demand for vector search capabilities is no longer a niche requirement; it’s a strategic necessity. We’ve seen this clearly in the market: a recent targeted community poll showed that 68% of enterprise leaders view integrated AI/vector workloads as a valuable future requirement, and a LinkedIn poll by our founder, Peter Zaitsev, showed that 73% of the community believes MySQL needs this functionality.
More importantly, we’ve received strong interest from strategic enterprise accounts and have seen the competitive risk firsthand as companies consider the future of the MySQL ecosystem.
Why Our Vector Solution Will Be Better
While several vector solutions exist in the MySQL ecosystem, each comes with significant trade-offs. Our goal is to deliver a solution that is superior for our users by being a native, fully open source, and true drop-in replacement for MySQL.
- Unlike DBaaS solutions, which lock you into a specific proprietary cloud platform, our solution will be fully open source, giving you the freedom to run anywhere.
- Unlike MySQL-compatible engines, which require a complex migration, our feature will be a simple, drop-in enhancement to the Percona Server for MySQL you already trust.
- Unlike basic plugins or connectors, which perpetuate the “two-system problem” we aim to solve, our native solution will provide the deep integration and transactional (ACID) guarantees that enterprises need for mission-critical applications.
Our plan is to deliver a solution that eliminates the complexity, cost, and data consistency headaches of managing a separate, specialized vector database. The initial version will focus on delivering native vector search and indexing, built on three core capabilities you need:
- Performant Search and Indexing: Create dedicated indexes on vector data and run fast, approximate nearest neighbor (ANN) searches.
- Seamless Integration with Standard SQL: Combine vector searches with the
WHEREclauses and business logic you use every day, all in a single query. - Full Transactional (ACID) Consistency: Ensure your vector data is as reliable and consistent as the rest of your transactional data.
Binlog Server: Reinforcing Operational Excellence at Scale
While we build for the future, we remain deeply committed to solving the immediate, real-world challenges our enterprise customers face today. For those running MySQL at scale, managing disaster recovery and distributed architectures remains a significant source of operational pain.
The demand for a better solution here is not speculative; it’s a direct and urgent request from large-scale enterprise users running complex, business-critical environments. They’ve told us they need a more robust and efficient way to manage their binary logs for Point-in-Time Recovery (PITR) and to scale their replication topologies without overloading their primary servers.
The Value We Plan to Build:
This isn’t just about archiving logs. Our vision is to create a foundational piece of infrastructure that makes your entire database environment more resilient and efficient.
The Binlog Server will be an active, online service focused on the core “Job to be Done”: enabling reliable, automated Point-in-Time Recovery. Key capabilities will include acting as a GTID-based live replication source, automated GTID state management, and enabling precise PITR workflows by timestamp or GTID set.
Where We Are Today & How You Can Help
We are firmly committed to building both of these high-impact features. Our immediate focus is on the initial capability validation process. We are working to validate a hyper-focused scope for each initiative to ensure we can deliver meaningful value to you as quickly as possible.
This is where you come in.
To ensure we build the right solutions, we need to validate our plans with the people who will use them every day. We are looking for interested users, developers, and DBAs to participate in our validation process. This is your opportunity to provide critical feedback on our proposed initial scope and help shape the future of these essential new tools for the MySQL ecosystem.
If you are building (or have plans to build) AI applications on MySQL or if you are facing the challenges of managing disaster recovery at scale, we want to hear from you! Please reach out to our MySQL Product Manager to participate in this critical next step. Book a 1-1 meeting here. Alternatively, please add your comments below to participate in the conversation.
Thank you for being part of the Percona community. Together, we will continue to build the future of MySQL.
