PgDog is funded and coming to a database near you (pgdog.dev)

Article summary

PgDog is a new open-source project that aims to make PostgreSQL horizontally scalable by adding a proxy layer in front of it. The project is funded and has already achieved 2M queries per second in production. PgDog is designed to work with any PostgreSQL deployment, including on-premises and cloud-based installations. The project's goal is to make PostgreSQL a viable option for large-scale applications without requiring significant changes to the database or application code.

Main themes

• PostgreSQL scalability
• Horizontal scaling
• Database proxy
• Open-source project
• Cloud deployment
• Database performance

What commenters say

• The performance gain of PgDog comes from its ability to load balance reads and scale out writes, rather than just being a rewrite in Rust.
• PgDog's architecture as a proxy layer allows it to scale the coordinator past a single machine, giving it an advantage over other sharding systems like Citus for OLTP workloads.
• Some users have experienced issues with PgDog's configuration and authentication, particularly in multi-tenant environments, and are looking for improvements in these areas.
• The use of connection poolers like PgDog is still necessary at large scale, even with modern databases, to manage thousands of client connections and distribute reads or shard to specific geographical areas.
• PgDog's comparison to other tools like pgbouncer shows that it can provide better performance and scalability, but some users have not seen a noticeable difference in their specific use cases.
• The lack of native support for secondary/primary writes and reads in PostgreSQL drivers makes it more difficult to use PgDog in scaled environments compared to databases like MongoDB.
• Some commenters believe that MongoDB's ease of use in scaled environments is due to its native support for topology changes and server selection, which PgDog does not currently match.
• Others argue that PgDog's open-source nature and ability to work with any PostgreSQL deployment make it a viable option for large-scale applications, despite some current limitations.